Sunday, January 31, 2021

The GameStop episode and efficient and interconnected markets

The GameStop (GME) episode shines attention again on some dark corners of finance. 

In short, for no particular reason a struggling physical retailer of video games attracted the attention of a Reddit forum WallStreetBets with 2 million subscribers. Investors pile on to the stock driving it crazily upwards. Incidentally, the same stock was shorted by some hedge funds, and their short positions may even have encouraged some of the Redditers. 

To give a scale of the trading bubble, the stock rise from $3.5 in August, 2020 to $18 on January 2, 2021 and then took off to touch $347.51 on January 27. The stock, with 69.7 million shares outstanding, generated short interest (or short positions due for delivery) of 71.2 million shares, and on January 22, 194 million shares were traded, which is over twelve times its average trading volume. At one point in time, more than 100% of GME stock was on loan.

This crazy run forced online brokerages like Robin Hood to restrict trading on GME and other bubble stocks as they ran out of funds to finance margin trades and faced pressure from their investors. Interestingly, it also put pressure on hedge funds with short positions on GME, with at least one, Melvin Capital, being forced into a rescue package from other investors.

There were at least four loops at work, each of which went beyond their useful limits. 

One, short-sellers bet against the stock by borrowing shares and selling short options. Sample this from a brilliant article by Matt Levine (HT: Ananth),

When you short a stock, you borrow shares and sell them, promising to return them later. You have to pay a fee to borrow shares, you have to post collateral based on the value of the borrowed shares, and you (generally) have to return the shares you borrowed if the lender asks for them back. When the stock goes up a lot, short sellers start feeling “squeezed”: Their borrow costs go up, they have to post more collateral, and lenders might asking for their stock back. Some short sellers might have to capitulate, and they will close their positions by buying back stock. There is a feedback loop: The stock goes up, short sellers give up, they buy stock to surrender, and their buying pushes the stock up more. 

And this graphic. 

Two, call buyers bet on the stock by buying options at a future strike price. This is great for small investors who want to bet on a rising stock. But if the stock price rises, the call option sellers are forced to buy more of the stock so as to cover their positions. This forces them into the 'gamma trap' of buying ever more quantities of the stock to limit their call losses. 

Three, brokerages like Robin Hood and ETrade provide margin loans to traders, which allow the latter to trade multiples of their capital. The brokerages place the trader's stocks as collateral and borrow to finance margin loans. This trade starts to unravel when lenders to the brokerages step back as the collateral becomes evident as bubble stocks like GameStop. This, in turn, makes brokerages raise their margins. Such margin calls on short-sellers force them to borrow ever more (or sell part of their securities). 

Finally, the brokerages themselves clear trades at the clearing houses (in the US, the Depository Trust & Clearing Corporation for equities and the Options Clearing Corporation for options), which stand between two sides of a deal by managing the risk to the market if one side defaults (a trade takes upto two days to be settled). The brokerages need capital at hand not only to finance margin traders, but also to pay for the margin requirements imposed by clearing houses as insurance against trades on their platforms, so as to protect its members against market volatility. As bubble stocks inflate and volumes balloon, clearing houses raise the margin requirements from their members, brokerages and investment banks, to ensure that deals are honoured. 

Finance 101 teaches us that the losses of option buyers are capped at their initial investment while potential profits are unlimited, and the losses of option sellers are unlimited. But a combination of short squeezes, gamma traps, and margin calls can destabilise entire markets.

In each case, perfectly fine strategies started to unwind. Two things were important - interconnectedness and speed of information (and misinformation) flows. Finance 101 considers both desirable attributes, in so far as they contribute to liquidity and efficiency in price discovery. But, as the GameStop episode shows, excessive liquidity and efficiency are dangerous.  

A confluence of often unrelated factors like Covid 19 WFH, ultra-low interest rates, emergence of zero commission trading platforms like Robin Hood, financing engineering like fractional trading, and discussion forums like Reddit involving largely day traders have brought stock market investing to the masses, but perhaps with undesirable consequences. Sample this,

Throw in social media Fomo (fear of missing out), Reddit’s mantra of Yolo (you only live once), and a bull market driven by Covid stimulus, and the GameStop saga looks inevitable.

And worryingly, all these are largely unregulated and un-coordinated actions of millions of retail investors in supposedly efficient markets. It questions the very premise behind markets as self-correcting.  

It was also a humbling moment for mighty hedge funds. They now realise that they could be at the mercy of the unforeseeable outcomes of an emergent collective.

To paraphrase Keynes, markets becoming more liquid and efficient increases the likelihood of  participants becoming insolvent!

Matt Levine has this description of the driving force behind such episodes,

The stock of a micro-cap company called Signal Advance Inc... shot up 5,100% after Elon Musk tweeted something about an unrelated app named Signal. The error, as it were, was quickly corrected: Lots of news stories, and a tweet from the “real” Signal, clarified that Musk was not talking about Signal Advance. The stock kept going up. (It’s still trading at roughly 10 times its pre-tweet price, weeks later.) Perhaps the buyers were impenetrably ignorant, but I suggested another possibility: There is a mass of retail buyers who like to all buy the same stock, and Musk’s tweet gave them a Schelling point to coordinate around. They weren’t confused about what Musk meant; they didn’t care that much about what Musk meant. They just like to all have fun together, pumping some stocks. You don’t actually need a Schelling point to coordinate around. You can just go on Reddit and talk about what stock you’re all going to buy... 

Bitcoin is a financial asset with no cash flows. It has value purely because people think it’s valuable. Bitcoin is worth $34,000 because other people will pay you $34,000 for it, and they’ll pay you $34,000 because other people will pay them $34,000, etc. There is no underlying claim; there is just a widespread acknowledgment that people think it’s valuable... It is an amazing collective accomplishment to create a new thing, from scratch, that is valuable just because we collectively agree that it’s valuable. It is amazing to find a way to create that collective agreement from nowhere. Once you have it, you can actually do useful things with Bitcoin—as a store of value, a currency, whatever—that you couldn’t do before. Bitcoin created real financial value out of, essentially, the human imagination. That’s cool but it’s also a terrifying proof of concept. If pure collective will can create a valuable financial asset, without any reference to cash flows or fundamentals, then all you need is a collective and some will. Just hop on Reddit and create value out of nothing. If it works for Bitcoin, why not … anything? Why not Dogecoin? Why not Signal Advance? Tesla Inc.? GameStop?

Or this by Jason Zweig,

Decades ago, small investors might pay as much as 5% to trade a stock. A stockbroker was a 9-to-5 guy in a paneled office who picked your pocket on every trade. Nowadays, your stockbroker is in your pocket, as apps on your phone let you trade stocks at zero commissions, anytime you want. WallStreetBets is the ultimate stage of this evolution. Thousands of people can amass small trades into giant pools of capital and whip each other into a collective frenzy. In what neuroscientists call “dynamic coupling,” the brain activations of different people doing the same task converge, firing in sync. In such situations, says Princeton University neuroscientist Uri Hasson, “I’m shaping the way you behave and you’re shaping the way I’m behaving. And coordinated behavior across many, many individuals can generate dynamics that are larger than anything they could produce separately.”


Thursday, January 28, 2021

An agenda for consultants in government

I have blogged earlier lamenting the alarming growth of the use of consultants in governments within India (and elsewhere) here and here. It's largely mediocre and routinised. 

I am no blind advocate of banning consultants within governments. But it's about using consultants for clearly defined tasks for which internal expertise is unavailable or external expertise is necessary. They range from preparation of specific documents to advising on specific projects or issues. It's the latter that is the subject of this post. 

How should consultant approach such work?

I am inclined to a two-step process. 

After a deep-dive on the issue, in the first stage, the consultant should distil their understanding into the different possible options available, with their respective pros and cons.

In the second stage, the consultant should exercise their considered judgement and make with a well-reasoned case, their recommendation about the preferred choice. This should focus on 

1. Why is the preferred choice superior to the alternative options?

2. What are the biggest drawbacks of the preferred option? Why is it still the preferred option?

The considerations in making this choice should include the technical merits, context (the implementation environment and bureaucratic capability to execute the same), comparative assessments (how others worldwide respond to the problem), prior experiences from similar contexts etc.

The two-step process is important for multiple reasons. The first step is important since the decision-makers within the government should know about all the available options and be able to make an informed choice. The second step is important since the government's choice should also take into consideration the opinion of professionally competent experts. Besides, it is also about holding the consultants too accountable for their advice. In fact, I would even suggest ranking consultants based on the success of their advice when adopted.

This thought discipline in government-decision making, enforced by the documents of record on the consultants' options and their preferred choice, is important for several reasons. For a start, it increases the likelihood of governments making the most informed choice. Another, it also forces accountability when the bureaucrats are overlooking the most qualified choice and making an alternative choice. It enhances the quality of the decision hygiene. 

It's a different matter that the government may, and rightly so, make a choice which is very different from that of the consultant. These choices can be dictated by legitimate and very valid real-world considerations like fairness and distributive justice, precedents, and political ideologies of democratically elected governments. Public policy choices are hardly only about expertise. 

I'll argue that it's therefore unfair and wrong for public commentators and academic scholars to critique governments just for not having adopted the preferred suggestion given by outsider experts. This point is important since it is the default media and popular commentary response when governments disregard the opinion of experts. Instead, their attention should be on the rationale that governments have for over-ruling the technical choice and making their alternative choice. 

This is also not to deny that governments do not make unjustifiable choices driven by nakedly political or corrupt considerations. That demands criticism. 

Governments should be willing to be held accountable for the reasons for their choices. Commentators should also be responsive to the right of democratically elected governments to make their fair political choices (for example, the choices that right and left wing governments make on various policies).

Granted that discerning between justifiable and unjustifiable choices by governments is difficult. But neither is it impossible.

Monday, January 25, 2021

Leave poverty alleviation to developing countries

Lant Pritchett reiterates his position on the primacy of economic growth.
Broad-based growth, defined as the process that raises median income, is far and away the most important source of poverty reduction. There is no instance of a country achieving a headcount poverty rate below 1/3 of its population (at moderate poverty line of $5.50) without achieving the median consumption of that of Mexico. This is not to say that there do not exist anti-poverty programs that are cost-effective and hence should be expanded, or, conversely, that there are anti-poverty programs that are not cost-effective (or even have zero impact on poverty) and should be cut back or eliminated. Analyses of these types of programs would enable a more efficient use of resources devoted to poverty reduction. But large and sustained improvements in global poverty will almost certainly have to focus on how to raise the productivity of the typical person in a poor country, which is a key source of national income growth.
And this from a review of an older book by Partha Dasgupta which tries to explain economics in terms of the lives of two ten year olds - Becky in a US midwestern suburb where her father works for a firm specializing in property law, and Desta in a village in southwestern Ethiopia, where her father farms half a hectare of land.
Perhaps the best that Becky’s world can do for Desta’s is to offer financial and technical assistance so as to promote and support local enterprises — including those involving education and primary health care — that people there are all too keen to create even as they see from a distance how people elsewhere have been able to improve their conditions of living. And perhaps the best that Desta’s world can do for Becky’s is to alert it to the enormous stresses economic growth there has put on Nature. There is, alas, no magic potion for bringing about economic progress in either world.
I am strongly inclined to argue that foreign aid should be confined to either development of pure physical infrastructure or for R&D or for state capacity building and should avoid advocating or supporting specific social development programs in areas like health, education, nutrition, agriculture etc. As I will try to explain, there is something about social development programs that demands that these societies struggle hard on their own to make difficult collective social and political choices.

If one looks at international development landscape and examine where the vast majority of thinking and attention is focussed, it is clear that poverty alleviation dominates economic growth.

It may be true that the share of spending on growth, in the form of aid assistance to develop infrastructure assets, is higher. But that is also because of the inherent capital intensive nature of those assets. Spending a billion dollars on a highway is several fold easier than spending a quarter of that to truly improve learning outcomes (not just construct school buildings).

As an illustration of the skewed priorities, how much of attention in today's mainstream international development is focused on improving tendering and contracting or third party quality audits for engineering works? How much is focused on addressing affordable housing or traffic congestion? How much on more effective utilisation of and equitable outcomes from irrigation systems? Instead, the attention on specific education and health programs or cash transfers or social protection is disproportionately large. 

Or take the example of the whole effort at outcomes-based financing and evidence-based policy making aimed predominantly at poverty alleviation. It is safe to say that these two phrases have become de rigueur in development circles, essential virtue signalling ingredients for donors and philanthropies to approve projects. They have gradually become the primary focus of economics research, both by academic researchers and by those in multilateral organisations and international think tanks. The program design agendas of multilateral and bilateral institutions and philanthropies are dictated by these two considerations.

It can appear pretty asinine to find fault with such apparently universal of ideas. Who would not want to focus on outcomes or base decision on evidence? That's a different story. I have critiqued the prevailing discourse on the former hereherehere, and here, and the latter hereherehere, and here with illustrative examples.

Back to external financing of development. What can be the main arguments against?

Externally advocated ideas like universal basic income, conditional cash transfers, a particular model of remedial instruction in schools, nudges, performance payments, community health workers, development impact bonds, technology solutions etc can have adverse unintended consequences in the long-run.

For sure, the practitioners and opinion makers in developing countries should become exposed to these ideas. They should be discussed and debated in the public domain. But they should make choices using their own resources. External funding, however well intentioned, cannot avoid unwittingly forcing choices on resource strapped countries.

These over-intellectualised and over-scholarized responses in the form of logically appealing and sanitised pilot evidence-based ideas and the financing to implement them prevents systems from facing up to the reality. It detracts societies and polities from traversing the path of hard and long-drawn collective struggles and making difficult trade-offs and attendant choices. It fails to account for the path dependencies and collective learnings required to succeed in addressing complex problems. It converts what are essentially social and political choices into technical fixes.

If Kenya needs to fix its school education system, its stakeholders need to grapple with the real reasons why children are not attending schools, why teachers are not accountable, why the quality of instruction is so poor, and prioritise resource allocation and make political choices accordingly.  There is a path dependency associated with reaching the destination. Technical solutions are a diversion from the real task.

For example, take the issue of teachers accountability to the parents. A biometric attendance solution is a good innovation but in a complex system can at best offer the illusion accountability and that too for a short-time, while also postponing the imperative to undertake the reforms like making the school and teachers accountable to the local community.

An idea like a Development Impact Bond to realise learning outcomes in education or change behaviours among prisoners in developing countries is mind-boggling in its ignorance. It is a seriously damaging digression from taking on real challenges. Like the PPPs, it peddles the irresistible illusion that governments need not bother about implementation challenges and can straight contract for outcomes, thereby generating value for money from public expenditures. In the context of concessions and Special Economic Zones (SEZs), Paul Romer famously postulated the test that a concession which cannot be gradually expanded beyond the SEZ to cover the entire country is inherently unsustainable. 

Or, a cash transfer program to overcome a state provisioning failure in housing or schooling or food security. It diverts attention and effort away from what is critical to enabling each of these in a sustainable manner, the hard and messy task of fixing state capacity. A cash transfer solution to each of these without fixing state capacity is like band-aid on gangrene.

World Bank and multilateral programs in areas like health and education are inherently biased towards getting money out of the door, with some capacity building trainings and governance related conditionalities attached whose compliance are most often perfunctory. This invariably gives a preference for physical infrastructure and personnel recruitment based initiatives, where pushing money out will be easy. Most often these programs are designed based on some standard universal template and pushed on to a system chronically deficient on multiple dimensions to implement them.

1. The main problem with external engagement in social development is that it distorts with ideas and money. The ideas are most often motivated by values, world-views, and concerns of those offering them, which are very different from those of the locals. External resources invariably prioritise interventions around these ideas. For all professions of high-sounding phrases like evidence-based policy making, outcomes-based financing, and systems thinking, the inherent biases and limitations of outsiders cannot be circumvented.

The problem is that this, especially because the ideas are also backed by aid money, ends up distorting the priorities of the developing country governments. The colonial hangover and the deeply entrenched faith that what comes from the likes of World Bank and Harvard Professors is "scientific medicine" for public policy, means that these ideas get embraced with limited resistance. For practitioners and opinion makers in developing countries adopting them becomes the default path for their countries to develop and is perceived as constituting progress.

2. Such entrenched inclinations effectively displaces all the struggles that the citizens of these countries have to undertake to diagnose their problems, internalise it, prioritise solutions and resources, and implement them. And by struggles, I mean everything from the public debates to political discussion, and focusing bureaucratic effort to prioritising financial resource utilisation. Countries need to go through the hard and long-drawn struggles and iteratively figure out what works best for them for a problem at this point in time.

Development is a long-drawn political process which involves struggles among interest groups, civil society organisations, political parties etc. Neither is there any technical short-circuit to this struggle nor is it desirable. In fact, limiting such struggles is likely to only create weak institutional scaffoldings. The works of Francis Fukuyama and Daron Acemoglu/James Robinson are just two exhibits. 

3. Another problem is that external engagements view social development as a self-enclosed and planned intervention to be implemented in finite time to realise definite end-state outcome. The objective invariably is to "solve" the problem. Unfortunately most social development "problems" defy this approach. These problems have no definitive "end" or "solution". Engagement with them is at best a long-drawn journey with uncertain paths and no definite destination, and involves work in multiple dimensions.

4. It also creates negative externalities. Such aid programs do not only suffer from the problems of isomorphic mimicry. Far more importantly, given scarce state capacity and political attention bandwidth, such programs also end up displacing implementation bandwidth from other more important measures.  Second, it allows politicians to get away with marginal tinkering by seen to be doing something new and big, instead of engaging with the real issues. Very valuable political attention bandwidth gets taken up by ideas that are mostly glossing over the real problems.

None of these are comments on the substantive merits of external ideas. In fact, many of them may actually be very promising. But the point is whether for this country at this point time (with all its problems and limited capabilities to address them), is prioritising this idea over others a good thing to be doing? The aim of this post is only to highlight the distortions that creep in when an appealing external idea gets introduced.

Like with anything, this too should not be taken to its extremes. I am not advocating autarky and isolation from ideas. Neither am I advocating a return to some old Luddite approaches. I am only suggesting that developing countries should keep their doors and windows open to ideas from outside but not be encumbered by them. They should undergo all the struggles required to prioritise their problems, figure out solutions, apportion scarce resources, and fight the implementation battles. They have to go through what Beckett said, "Ever try, ever fail, try again, fail again, fail better".

Isn't it telling something profound that the biggest development stories of our times, East Asia and China, have been without any of these new ideas and have been built on equitable access to opportunities, pragmatic governance, and plain simple economic growth, ingredients that every one of today's developed countries too had in their growth phases? Developing countries should read Joe Studwell.

This verse from the Chapter 6 of Bhagavad Gita applies just as much to countries as it does to people. Lift your selves up your Self!

Saturday, January 23, 2021

Weekend reading links

1. Janan Ganesh makes the point brilliantly about today's television news,

The Federal Communications Commission scrapped the “fairness doctrine”, freeing US broadcasters to editorialise. Throw in some cable upstarts and you have the bagginess, the histrionics, of today’s TV, where all news is BREAKING, and each anchor a monologue merchant. It is hard to convince the young how terse and inscrutable these desk-bound enigmas once were. The debasement of TV has done far more than social media to pollute civic life... What I have in mind is a subtler loss of rigour, a widening of what counts as “news”, even among those with no agenda. Reporting can still be heroically brave. But what used to be a bone-dry recitation of the facts is too often padded out with interpretation, contention and the quest for a story’s “meaning” (leave that nonsense to columnists). Were the phrase not taken by another medium, we might call it Impressionism.

2. FT on the rise of the private debt market, which rose from $575 bn at end of 206 to $887 bn by June 2020.

3. In the waning days of the Trump Presidency, the National Security Council has declassified and released a 2017 document called the US Strategic Framework for the Indo-Pacific, which forms the basis for US engagement in Asia. 

The "desired end states" of the policy takes a very favourable view of India,
India's preferred partner on security issues is the United States. The two cooperate to preserve maritime security and counter Chinese influence in South and Southeast Asia and other regions of mutual concern. India maintains the capacity to counter border provocations by China. India remains preeminent in South Asia and takes the leading role in maintaining Indian Ocean security, increases engagement with Southeast Asia, and expands its economic, defence and diplomatic cooperation with other US allies and partners in the region. 

4. Bloomberg asks three Wall Street executives about the biggest risks they foresee to the world economy and society. Martin Chavez talks about the increasing digitisation and interconnectedness of the global financial flows and its risks. Eileen Murray talks about the risk of mass unemployment due to automation and its consequences on the society and polity. David Siegel talks about how modern technologies could end up devaluing human beings. 

5. Very informative summary on India's banking sector over the last two decades by Tamal Bandopadhyay. Interesting that the banking sector average deposit and credit CAGR declined in the two decades from 17.8% to 11.7% and from 22% to 11.2% respectively. 

6. The lenders to DHFL will take away only 33% of the total debt worth Rs 90000 Cr from the proposal by the Piramal Group which has been approved by the Committee of Creditors. The shareholders will be wiped out.

7. The Financial Express reports that 18 GW of solar projects with aggregate investments of nearly Rs 1 trillion may be facing problems as discoms are reluctant to sign PPAs. The Solar Energy Corporation of India (SECI) is the aggregator who signs PPAs and then markets power supply agreements (PSAs) with discoms. The problem is the rapidly declining tariffs in succeeding auctions, which forces discoms into waiting out for better prices or even reneging on their existing PPAs. 

It may be time for SECI to abandon its current build it and they'll come approach and adopt a market-building approach by firming up a pipeline of PSAs and then signing PPAs. It is also very important that supply from all PPAs during some period, say, a year, are bundled together and marketed on a flat rate to the discoms.

8. After buying sporting teams/clubs, private equity is now eying stakes in governing bodies that run competitions. FT reports that Silver Lake is taking a 15% stake in a new entity that would hold NZ Rugby's (the governing body which runs the rugby union team and manages competitions) $2 bn broadcasting, sponsorship, and ticketing rights. FT writes,

Silver Lake acquired a $500m stake in the parent company of English Premier League football club Manchester City in 2019 and is an investor in Ultimate Fighting Championship, the mixed martial arts competition. But the pioneer in private equity investing in sporting tournaments has been CVC Capital Partners, a Luxembourg-based buyout group that previously owned Formula One and MotoGP and is in talks alongside fellow buyout group Advent International over a €1.6bn investment in Serie A, Italy’s top football league. CVC has triggered the recent scrum for rugby deals. It has spent the past two years hoovering up stakes in club contests such as the English Premiership and Pro14, and is in the final stages of acquiring a £300m share in the Six Nations, Europe’s leading national team tournament.

9. Business Standard has an article on the deal making and global transportation of the Covid 19 vaccine.

10. Jonathan Ford has a good article on the consequences of the debt binge by local governments in UK. As rates stayed low and government transfers declined, local governments sought to augment their incomes by taking loans to invest in real estate and other commercial ventures. 

Local authorities spent £6.6bn on real estate between 2016-19, according to the UK’s National Audit Office — 14 times more than in the previous three-year period. But they also bet on other commercial enterprises, from renewable energy producers to start-up banks. The idea was to bolster authorities’ incomes with earnings from real investments, helping to sustain austerity-hit public services such as libraries and schools. A sort of pro-social carry trade. The risks councils are running are eye-watering. Take Spelthorne, a small authority near London, with a core annual budget of around £11m. It borrowed more than £1bn to build a property portfolio mainly outside its own locality. Or Thurrock in Essex, which has similarly borrowed £1bn (almost five times its £220m budget), much of it then invested in unnamed renewable energy schemes.

11. The Economist writes about the rising importance of microchips and the geopolitical struggles around controlling its production and supply chain.

In the 20th century the world’s biggest economic choke-point involved oil being shipped through the Strait of Hormuz. Soon it will be silicon etched in a few technology parks in South Korea and Taiwan... A surge in demand and those novel kinds of computation have led to a golden age in chip design. Nvidia, which creates chips for gaming and artificial intelligence, is now America’s most valuable chip firm, worth over $320bn. The quest to create bespoke chips in order to eke out more performance—think less heat, or more speed or battery life... In the future a new open-source approach to designing chips, called risc-v, could lead to more innovation... Moore’s law, which holds that the cost of computer power will fall by half every 18 months to two years, is beginning to fail. Each generation of chips is technically harder to make than the last and, owing to the surging cost of building factories, the stakes have got bigger. The number of manufacturers at the industry’s cutting-edge has fallen from over 25 in 2000 to three. The most famous of that trio, Intel, is in trouble... It may retreat from making the most advanced chips, known as the three-nanometre generation, and outsource more production, like almost everyone else. That would leave two firms with the stomach for it: Samsung in South Korea and TSMC in Taiwan... An array of corporate A-listers from Apple and Amazon to Toyota and Tesla rely on this duo of chipmakers.

The other big industry rupture is taking place in China... The American tech embargo... is starting to bite. In the last quarter of 2020 TSMC's sales to Chinese clients dropped by 72%. In response, China is shifting its state-capitalist machine into its highest gear in order to become self-sufficient in chips faster... A $100bn-plus subsidy kitty is being spent freely: last year over 50,000 firms registered that their business was related to chips—and thus eligible... The manufacturing duopoly could start to use their pricing power. Already a fifth of all chip manufacturing, and perhaps half of cutting-edge capacity, is in Taiwan, which China claims as its own territory and threatens to invade. The chip industry is poised for mutually assured disruption, in which America and China each have the ability to short-circuit the other’s economy... chip-users such as Apple should press TSMC and Samsung to diversify where factories are. America must urge Taiwan and South Korea to cut their soft subsidies for chip plants, so their firms have more incentive to build factories around the world.

12. Finally, a very good article on James Simmons who just stepped down as Chairman of Rennaissance Technologies, described as the world most successful hedge fund. 

From 1988 through 2018, Renaissance’s flagship Medallion Fund had an average annual return of about 40% after fees, with almost no money-losing years; before fees, its returns were even more eye-popping. Although Medallion’s trading strategies do eventually get discovered by the market, forcing the company to find and exploit new inefficiencies, their profitability tends to last for decades rather than mere months. That steady outperformance has earned Simons an estimated fortune of $22.9 billion, according to the Bloomberg Billionaires Index.

The article points to the limitations of the Renaissance model,

The Medallion Fund’s returns don’t compound very much. Typically, we think of financial returns as compounding exponentially over time. If you invest $1,000 in the S&P 500 Index and it grows at a 10% annualized rate over 40 years, you’ll end up with about $45,000. But Medallion’s 40% returns can’t work like that. The fund currently manages about $10 billion. At a 40% annual rate of return over 20 years, $10 billion would turn into $8.4 trillion — more than twice the amount currently managed by all the hedge funds on the planet. 

Obviously this doesn’t happen. The Medallion Fund makes regular profit distributions so that the total amount of money in the fund stays at just a few billion dollars. Investors aren’t allowed to keep all their money in Medallion year after year until it goes to the moon; instead, they can only keep a portion there, earning money steadily but not exponentially. Undoubtedly, the reason Medallion stays relatively small is that its strategies don’t scale up. The market inefficiencies that allow brilliant mathematicians to make steady money aren’t infinite; bet enough money on these strategies, and they stop making huge profits, as prices move into a more efficient alignment. This is why Renaissance doesn’t let investors keep all their money in Medallion, and no longer solicits money from new outside investors — more money would drive down returns. Renaissance does operate a number of other funds, which use different strategies that scale up better, and are available to outside investors. But these more capacious funds necessarily lack Medallion’s special sauce. During the topsy-turvy year of 2020, as Covid-19 upended financial markets, Medallion reportedly racked up a 76% return. But three of Renaissance’s largest funds lost big money in 2020.

Thursday, January 21, 2021

The illusory appeal of bad banks

The RBI's Financial Stability Report estimates the gross non-performing assets of the country's banks to reach 13.5% by September 2021. It comes amidst discussions about a bad bank. The Business Standard has an editorial which dwells on this topic. 

Fundamentally a bad bank idea has three parts:

1. Establishment of a bad bank and freeing up of the resultant good banks to resume normal lending activity.

2. Transfer of bad assets from all the banks to the loan book of the good bank.  

3. Resolution of the bad assets held by the bad bank and maximise their recovery. 

While hardly trivial, the first two steps are nevertheless only administrative exercises. Here too things like valuations while transferring assets can be tricky. The banks will have the incentive to maximise their valuations. But given these are G2G transfers, this is a problem that can be surmounted. 

But it is the third step that is where the real challenge lies. And without significant progress on the third step, the first two steps are costly diversions. The net cost on the system (markets and the government) would remain the same, perhaps become even higher. 

The question that then needs to be asked is this. What's the binding constraint on the resolution of bad assets? Are the bad assets amenable to the standard resolution and liquidation processes? Is it lack of specialised expertise and adequate management time, that the regular banks could not have done it on their own? 

I believe that a bad bank solution's popular appeal comes from its proximate effect of being able to separate the good and bad assets, free up the balance sheets of the regular banks, and therefore restore normalcy in the credit markets. This is illusory and like sweeping things under the carpet or kicking the can down the road. It overlooks the reality of having to resolve the assets transferred to the bad bank at some fair enough price, so as to complete the process. 

Most commentators, including eminent academic researchers, are carried away by the proximate steps and first-order effects in their prescriptions to complex public policy challenges. Further, as this BIS paper finds, bad banks also require a complementary set of steps for success

The main finding is that bad banks are effective in cleaning up balance sheets and promoting bank lending only if asset segregation is combined with recapitalisation of the bank's balance sheet. Used in isolation, neither tool will suffice to spur lending and reduce future NPLs. Looking at a wide range of episodes, we find that assets segregation is more effective when (i) asset purchases are funded privately; (ii) smaller shares of the originating bank's assets are segregated; and (iii) asset segregation occurs in countries with more efficient legal systems.

The Indian context may struggle to meet many of these requirements.  

The resolution of bad assets has several dimensions. As I have blogged here, the biggest issue in India, unlike the western countries where bad bank ideas have been tried, is that the banking system here is largely publicly owned. This has at least four implications. 

One, given their size, the large inevitable haircuts to be taken during the resolution are fiscal costs to be borne by the government. This by itself is a significant issue. Second, the officials in the resolving bad bank face the acute problem of stifling and paralysing external oversight on their actions from investigative and prosecuting agencies and the judiciary. The judgements of all these agencies are most often based on limited understanding of the context and the business activity as well as deeply hindsight biased. 

Then there is the issue of resolvability of these assets. As I have blogged earlier, the resolvability of largely public good infrastructure and related physical assets, both operational and at different stages of completion, is qualitatively different when compared to the resolution of standard mortgages, consumer retail and credit card loans, corporate loans, student loans etc. The bad banks in the west were deployed to resolve largely the former category of assets. The former requires a more heterodox approach tor resolution. 

Finally, there is the issue of the market's ability to absorb these volumes of bad assets in one flush. The hyper-ventilation by commentators on resolution overlooks the limited depth of local capital and relatively very small foreign capital interested in such assets at fair prices. There would obviously be willing buyers at exorbitant haircuts. Who would not want assets at a steal? But would the political economy and the fiscal burden allow for such resolutions at scale?

In view of all the aforementioned, with bad bank we would not have moved much on the issue of putting to bed the problem of bad loans. We would only have transferred the problem from one finger of the government to another. Besides it would have come at a high cost in terms of the bandwidth it would occupy among all stakeholders concerned, diverting effort within banks from pursuing their main activity for at least six months.  

The solution instead is to retain the assets on the bank's balance sheets and resolve them through a heterodox approach as outlined here, here and here that depends on the nature of the impaired assets. 

One, continue the ARC and leveraged buyout fund routes. Two, encourage purchases by private (infrastructure funds) and public (like NIIF, IIFCL etc) funds which could either securitize some assets or establish SPVs and manage these assets till they become profitable. Three, strategic acquisitions of some of the infrastructure assets, especially in power, by the better governed public sector units. Four, strategic auction of certain other assets, especially in steel and metals, to reputed private buyers. In at least some of these cases, provisions like back-ended clawback of some share of windfall gains by potential buyers may ease the resistance and apprehensions associated with such sales.

Even this is not going to be easy. But it stands a far greater likelihood of success than the illusion of bad bank. 

Wednesday, January 20, 2021

Simple management productivity improvements and state capability

Often, the most impactful things to address a problem are the simplest and most obvious ones. But, for a variety of reasons to do with cognitive biases, people generally overlook them and instead explore more difficult and innovative things and invariably end up falling short. 

This has relevance in debates to improve state capability. Most of the search for state capability improvements revolves around technology interventions and e-governance, process re-engineering, additional manpower, new institutional arrangements and so on. A simple and obvious area of better management of administration is often overlooked.

Consider the following questions and the areas of administrative management they highlight. 

Given the vast responsibilities and scarce (capable) personnel and resources, what should be the prioritisation of work by officials? What is the basis for this prioritisation? Is it aligned toward maximising realisation of organisational objectives?

Is work-allocation among different officials within the agency optimal? Does it take into account the realities of vacancies and the preponderance of recalcitrant and inefficient officials? 

What is monitored by officials at each level? Is it right-sized, in terms of neither too much nor too little? What is the periodicity of monitoring? How is the monitoring done at each level? What is the mechanism to ensure adherence to monitoring protocols and assess the quality of monitoring?

Are meetings organised most effectively - in terms of their periodicity, whether clear and brief agendas are communicated in advance, what gets discussed, and how the minutes are recorded? How are the meeting outcomes followed-up? How are failures to comply addressed?

How do managers manage their staff? What are the thumb rules followed to extract work from the different categories of officials (by capacity and attitudes)? How is managerial accountability enforced? 

I'll venture to argue that the typical administrative unit and its bureaucracy in a developing country is entrapped in a very low-level equilibrium. In the circumstances, in line with the objective of moving from poor to average or good, even a satisfactory performance on each of these can have dramatic effect on the entire system, one which is far higher than that realistically possible with the more commonly pursued interventions and innovations. In most systems, the prevailing status is most likely so poor as to leave us with low-hanging fruits in terms of potential productivity improvements. 

All that is required is a leader who is committed enough to exert some basic level of administrative oversight and aware enough to pursue some basic management techniques and impose the same across the organisation. It is for this reason that it's common place to see hitherto moribund systems or organisations becoming abruptly galvanised by the arrival of a leader with a high level of commitment and professional competence. And also relapsing back to its original state after that person gets transferred. 

While waiting for such a leader is not an institutional solution, it's a pointer to prioritising the adoption of basic management practices. This is about the adoption of very simple and basic work, people, and situations management techniques, and not the sort of stuff one learns from management schools. Unfortunately, it's not an area that receives any attention in conventional academic research and management consulting.

Monday, January 18, 2021

Market access and concentration in digital markets

One of the most important arguments against greater regulation of the technology companies is the view that despite their large market shares they enhance consumer welfare. And since consumer welfare is the primary objective of market regulation, there should not be any regulation of these companies. 

This is a view that can be traced to both the ubiquity in our daily lives of the convenience value of these innovations and the recency of these markets. We routinely engage with all of these innovations - e-commerce, social media, search engines, streaming video, fintech etc - in our daily lives. The have enriched our lives in some way or the other, and continue to do so. 

The relatively early stages of their market evolution also mean that their benefits stand out more than their costs. In other words, there is a representativeness bias towards their benefits over their costs. But it's only a matter of time before the benefits recede in our individual and collective cognitive selves and their gradually manifesting costs become increasingly salient.   

Regulators and judges are no less immune from these biases and influences. 

In this context, if regulators across the world were looking for exhibits on the enormous and potentially dangerous control that the big technology companies exercise in their markets, then they need to look no further than the ejections of @RealDonaldTrump from Twitter and Facebook, Parler from Amazon Web Services, Fortnite from Apple Pay, and PayTm from Google Play. 

There are two distinguishing features in all these cases - the consequences of disproportionate levels of market access and market concentration. The first concerns the balance of market power and the unilateral nature of the decisions by the technology companies, and the second is the effective loss of market access for each of those ejected. 

A logical extension of the first confers on the technology companies a vastly disproportionate power to turn off the market access tap on any user of its services. This is unprecedented since unlike regulated markets like infrastructure services, there are no proximate regulators in these markets. The legal recourse is most often long enough to effectively eliminate the aggrieved user (or company). 

The second is a demonstration of the massive market power that these companies enjoy. It is true that these technology companies pioneered the creation of their respective markets. But those markets have come into being, it is only appropriate that it be subjected to the same set of rules that govern all other markets. The platform nature of the markets only means that the original market has since spawned several other inter-connected markets, thereby making market concentration even more dangerous. 

These are powers that previously only governments had. As a comparator, if you consider internet as the highways system, it's like a large vehicle manufacturer having the power to both prohibit someone from using the road and also make competing vehicle manufacturers unattractive (say, because of inadequate servicing options) for users. Or, if you compare internet advertising with with television advertising, it's like the cable operator controlling what advertisements can be displayed and what content can be seen on your television.

The growing pile of high-profile and market disrupting ejections may well be the triggers that regulators and judges need to tip over their cognitive selves from being dazzled by the convenience of these innovations and into recognising the costs and dangers of market concentration.  

The unambiguous takeaway, one that will become increasingly evident in the coming years, is that market concentration on the one hand and market competition and consumer choice on the other cannot co-exist. They are the impossible dilemma. Worse still, in the absence of competition, innovation and economy-wide dynamism suffers. This has been historically acknowledged for all markets, and digital markets are no different. Perhaps even more so for them given the network effects and their platform nature, both absent in case of regular markets. 

I have blogged earlier here (beginning of anti-trust actions in US and Europe), here (Google and anti-trust challenge), here (anti-trust challenge in US), here (market monopolisation), and here (Amazon's market abuse of startups) on the problems with market concentration and the need for regulation. This points to what Google founders themselves though about data monetisation and digital advertising. I have blogged here, here, and here on the problems with regulatory arbitrage and here on the market service quality problems due to limited and poor regulation. And we are not even talking about the several other distortions arising from such market concentration, especially the valuation bubbles in the financial markets - see this and this

Finally, this is a summary examining the dynamics of digital markets and how it offers a different perspective on these markets, and this is a summary of how the big technology companies have become digital gatekeepers to large and critical markets. 

Update 1 (23.01.2021)

The Australian legislators have opened up a new front in the battle against Big Tech with a new law that would compel social media groups like Google and Facebook to pay news organisations and publishers in exchange for circulating stories. If Google is making money from advertising on a page with a link to New York Times which is what is attracting people to that page, why should not NYT demand a share of the advertising revenue that Google attracts due to positioning the NYT link?

Saturday, January 16, 2021

Weekend reading links

1. Several interesting insights about India's labour market trends from Mahesh Vyas,

Urban India accounted for 32 per cent of total employment in 2019-20 but it accounted for 34 per cent of the total loss of employment in 2020-21 till December 2020... Women account for a mere 11 per cent in total employment. But, they accounted for 52 per cent of the job losses... All age groups below the age of 40 suffered a fall in employment till December 2020 this year while all age groups above 40 years of age have seen a small gain in employment... while twenty per cent of the working age population is in their twenties, they account for a lower 19 per cent of the employed persons. Worse still, these twenty-somethings account for 80 per cent of the job losses as of December 2020. Folks in their thirties account for 17 per cent of the working age population. They account for a higher 23 per cent of the total workforce... but they accounted for 48 per cent of jobs lost as of December 2020. There were job gains in senior age groups – those over 40 years of age. Evidently, India’s workforce aged in 2020-21 during the lockdown. The share of those over 40 years of age, which was 56 per cent in 2019-20 increased to 60 per cent by December 2020... Graduates and post-graduates had a 13 per cent share in total employment in 2019-20. Their share in the loss of jobs was 65 per cent. Of the 14.7 million jobs lost, 9.5 million were those of graduates and post-graduates. Finally, salaried employees who accounted for 21 per cent total employment in 2019-20, accounted for 71 per cent of the total job losses.

2. As he leaves office the balance sheet of President Trump's trade actions in terms of US trade balance against China is not flattering.

3. Hugo Erken and Michael Every of Rabobank argue that India was wise not to join RCEP. Their central argument is that since most of RCEP members are net exporters, their major share of trade is external to the bloc whereas it forms 70% of India's trade, India has among the highest average tariffs and lowest non-tariff barriers (NTBs) while it's the reverse in case of the major RCEP members, and since RCEP does not cover NTBs, India entering RCEP would make it something like a buyer of last resort, thereby worsening its trade balance and hurting domestic businesses.

4. Morgan Housel on the power of compounding,
$80.7 billion of Warren Buffett’s $81 billion net worth was accumulated after his 50th birthday. Seventy-eight billion of the $81 billion came after he qualified for Social Security, in his mid-60s... 99% of Warren Buffett’s net worth came after his 50th birthday, and 97% came after he turned 65.

5. Damien Ma, Houze Song, and Neil Thomas at Macro Polo have a report on Global Value Chains. It uses the examples of Li-ion batteries, OLED displays, and AI chips to discuss GVCs. This about the Li-ion supply chain,

These raw materials and the main components of the battery—cathode, anode, separator, electrolyte—constitute the upstream and midstream segments of the supply chain. For inputs such as lithium, cobalt, and graphite, countries like Chile, Argentina, DRC, China, and Australia dominate. For the key components of the battery in the midstream, that supply chain is dominated by Japan, South Korea, and China, with the US playing a role in supplying separators. In terms of downstream production of battery cells, China commands a 61% global share, while the US share of 9.5% is centered almost entirely on a single company: Tesla.

On OLED displays that are a feature of smartphones,

An important component in any OLED screen is glass, particularly the specialty Gorilla Glass that is manufactured by US-based Corning... Beyond the cover glass, other types of specialty glass are used in the various layers of the OLED display such as the encapsulation. Other layers of the display include the front plane, which is the emissive layer composed of organic materials, and the backplane that includes the IC driver and TFT. These different layers constitute the upstream and midstream of the supply chain. The main raw material for glass is silica sand, which is abundant around the world, with the United States (40%) and China (27.5%) being the two leading producers. The midstream—composed of the display glass, IC drivers, and OLED materials—is dominated by South Korea, Japan, and the United States. When it comes to display glass, the United States and Japan are basically at about 48% and 47% global market share, respectively. America is also a leading supplier of OLED materials, with roughly a 45% market share, while South Korea and Taiwan combined make up some 80% market share of IC drivers. Finally, South Korea’s Samsung dominates the downstream OLED display production, accounting for 90% of the global end-use production capacity.

And on AI chips, or customised chips that support the computational demands of AI technologies, 

Most chips are built on top of core underlying architectures whose intellectual property (IP) is held by a handful of key firms... Manufacturing a chip requires core IP, design, fabrication, and assembly, which together form the midstream segment of the supply chain. Japan, the United States, and the United Kingdom dominate the chip IP segment, with UK’s ARM holding nearly a 45% market share. China so far has one company that is competitive in this arena, but it only holds a 1.8% market share. The United States, the Netherlands, and Japan dominate the crucial market for semiconductor manufacturing equipment used in fabrication. Dutch company ASML virtually monopolizes the supply of photolithography machines, which each cost around $120 million. Taiwan plays a crucial role in the actual fabrication of chips. Taiwan’s chip foundries make up more than 80% of the global market, while mainland China’s SMIC foundry has just 6.6%. But China is gaining ground on Taiwanese firms when it comes to the final stage of chip assembly, which is typically handled by low-margin outsourced contractors. The downstream segment, which is determined by global sales of semiconductors, is dominated by the United States, South Korea, and Japan. In fact, no Chinese company yet appears in the top ten in terms of global sales. American companies occupy 35% of this market, while Korean, Japanese, and Taiwanese companies combined make up roughly one-third.

The global interdependencies in all the three are very clear, making disruptions in one place a matter of global concern. 

This has excellent graphical data about all the three industries.

6. Contrary to opinions that question the wisdom of establishing a well-capitalised Development Finance Institution for infrastructure funding, there is a real need for such an institution. And it needs to be state-owned to make any meaningful difference. Even with the inevitable political economy risks, if infrastructure financing has to be mobilised, there are few other alternatives. This long paper makes a very detailed case.

7. The Economist has a briefing on innovation and public support for it. This about how the post-war years were characterised by technologies that people and firms used,

The post-second-world-war years were not only marked by a growth in government r&d spending, but also by the scientific excellence of in-house laboratories at companies such as at&t and ibm. In the 1960s researchers at DuPont published more articles in the Journal of the American Chemical Society, the field’s leading journal, than mit and Caltech combined. The production of scientific knowledge and the desire to solve real-world commercial problems were closely entwined. Science was being pulled into the economy, not just pushed; this was the environment in which, in the 1960s, the term r&d was invented... corporate science has gone into decline, with big firms increasingly choosing to license research from universities rather than do it themselves. Further removed from production, the universities which serve as the primary research focus in many countries are not so focused on useful invention. If the current innovation system is simply less good at creating growth-boosting innovations than it was, then spending more on r&d will not raise incomes as much as it might. It may simply produce more research papers.

This is a claim which deserves much greater scrutiny,

Amazon claimed to spend $36bn on “technology and content” last year, more than the science budgets of Britain and France combined.

I struggle on this. Spending $36 bn to tweak algorithms and software and doing data analytics??? Is this more of accounting gymnastics?

8. Gillian Tett points to signatures of return of inflation,

What was the best-performing asset class in 2020? If you think “tech stocks” or “bitcoin”, think again. Instead, as the Bridgewater hedge fund recently wrote to its clients, “among the more interesting and least recognised outcomes” of 2020 was that US inflation-linked bonds beat other assets by delivering a 35 per cent return, on a risk-adjusted basis, as investors hedged against inflation risks.

9. India's credit growth decline problem in a graphic

Thursday, January 14, 2021

Summary of posts on understanding and doing public policy

I thought of summarising all of what I consider to be useful posts on public policy, especially for commentators and researchers who engage on these issues. 

While most insiders in government take many of these for granted (though not always with a full understanding of its implications for policy making and implementation), very few outsiders appear to be even aware of these as issues. It manifests in the constant articulation of logically great but impractical ideas in public commentary as well as in the anger and frustration at governments in general. Consultants, researchers, and general opinion makers are particularly prone. 

Most of these posts should be seen as merely surfacing very important but less discussed aspects of development and policy making. Where solutions have been suggested, they are only illustrative pointers. 

So here goes, for whatever it is worth, the list of posts:

On policy making and implementation

1. Importance of trust and delegation for public official managing large organisations or programs so as to prioritise their work on the handful of issues which are truly important and need attention and effort. This on the role of simple management practices in improving state capabilities. 

2. This on the two imprimaturs of sophistication and professionalism in our times, the application of the scientific method to development, the rise of technocracy - the objectivity of quantitative methods, and the wisdom of experts.

3. This, this, this, this, and this on the perils with pursuing a logically appealing strategy of purely outcomes-based targeting of development objectives. It is very counter-intuitive.

4. This, this, and this about the problems with the use of performance incentives in public services.

5. This (Aadhaar and internet access), this (banking sector supervision), this (use of technology in targeting), this (health care), and this (DBT transfers) about the limits of digitisation and applications of technology in solving persistent development problems. They also highlight the importance of retaining traditional physical inspections based oversight. 

6. This and this about the problems with the elimination of middlemen.

7. This and this about the problems of excessive reliance on aspirational standardisation in development policy making and this about the trend of harmonisation of trade and other practices motivated by imperatives of globalisation. 

8. This and this on the problems with the disproportionate focus on efficiency seeking interventions at the cost of resilience etc in businesses and governments which end up detracting from realisation of the objectives. 

9. This on the need to shift the paradigm on public service delivery away from efficiency toward quality. 

10. This on the reality that supply side is absent or severely constrained to substitute or outsource to private providers the many services offered by governments. 

11. This, this, and this on the problems with the excessive application of financial incentives to attain development goals. 

12. On wicked problems, this and this from agriculture.

13 This and this on the general problem of excessive reliance on experts and technocracy. Some examples - from Covid 19 response, and this and this from central banking,

14. On the importance of the small details of implementation - from agriculture, factoring receivables, this and this with labour market, export promotion schemes, and MUDRA scheme. 

The gist of some of these are compiled in this collection on behavioural challenges for IAS officers co-authored with Dr TV Somanathan.

On International Development

15. This and this on impact investing. 

16. A reality check on some of the excessively hyped areas of development innovation in impact investing - financial engineering, fintech here and here, micro pensions, crop insurance, and micro-insurance

17. On the problems with the excessive reliance on consultants in development here, here, here and here. This agenda for consultants to governments.

18. On problems in the academia - conflicts of interest, peddling false narratives, claiming more than there is from their research etc. 

19. On problems with specific RCTs - impact of Aadhaar on welfare programs delivery, hotspot policing, third party quality audits in pollution control, telephone feedback solicitation, policing reforms.

20. On general problems with RCTs here (implementation validity), here (displacing other research on India), here (scale validity challenge), here, here, and here

21. On research-reality challenges - trade-off between academic rigour and relevance, dataisation and ahistoricism. This and this about the limitations of the Raj Chetty school of data economics research in its relevance for policy making.

22. On general problems in development thinking - enduring fallacies, discourse of development, experience discount, inadequate context appreciation (Jensen paper) etc.

23. This and this on the excessive reliance on smartness in solving development problems.  

24. On the deeply misleading campaign surrounding evidence-based policy making - this on evidence generation to satisfy the urges of researchers and outsiders, this and this on the need to adopt a non-ideological approach and prioritise the use of evidence arising from latent institutional knowledge, and this on why policy makers find the vast pile of evidence generated largely useless. 

25. This, this, and this on the smart-outsider-knows-best problem in development.

26. Finally, this on why development is a faith-based activity. 

Monday, January 11, 2021

More on growth promoting and hindering corruption

I had blogged here earlier about Yuen Yuen Ang's latet book, China's Gilded Age: The Paradox of Economic Growth and Vast CorruptionThis podcast is a good summary of the book. And this and this book discussions too. 

She classifies corruption using a framework as below. I had blogged here earlier about this framework. 

She makes the distinction between growth damaging and promoting forms of corruption. Theft like extortion and embezzlement belongs to the former and access money to the latter. She defines access money as the purchase of influence by the capitalists from those in power. Access money exacerbates inequality. In this model, state-business relationship are not extractive, but transactional. 

She uses the framework to assess corruption in different countries.
In countries like India power is diffused across large and fragmented bureaucracies leaving officials with limited power to get things done on their own but with significant power to create obstacles, bribes are primarily paid to overcome the numerous hurdles created by the bureaucracy. In contrast, in China since power is concentrated in local leaders who have the power to definitively confer favours, bribes are made to win those favours.
Yuen's central point is that Chinese growth has been powered by access money, whereas the growth hampering forms of theft have declined. 
Yuen describes China as a "capitalist dictatorships disguised as communist", with the dominant belief of each according to his needs, to each according to his ability and connections. Consistent with this, she also describes China as a "corrupt meritocracy". 

But this, she argues, in turn, creates its set of distortions and risks. Xi Jinping thinks corruption is an existential problem, in so far as it can destroy both the Party and economic growth. He is worried about grand transactional corruption enmeshed with patron-client relationships among political elites and senior party leaders. 

Yuen points to two fault lines on how corruption can impact the Chinese political system.

Chinese corruption can intensify factional rivalries. Fiefdoms amass astronomical rents and become  ambitious and defiant of the top leadership. They also become powerful enough to take on central leadership when their interests are hurt. Such internal implosions are not non-trivial.

Another rupture point is when the structural risks linked to crony capitalism implode. For example the debt pile has the potential to deeply destabilise the system. This also explains Xi's focus on debt reduction and deflating the property bubble. 

Another risk is that of corruption moving away from the traditional real estate, infrastructure, and manufacturing sectors. Take the example of Government Guiding Funds (GGF). As of end-2017, there were 1500 of them managing $1.4 trillion in funds, with the mandate of investing in technology and innovation sectors. It is a form of industrial policy that uses public funds as seed money to invest in these sectors. The GGFs are effectively funds of funds, allocating the funds to traditional venture capital and other investors to in turn invest in innovation and technologies. This could become the new source of corruption. It is not clear how these funds are disbursed, to whom, and why. Holding fund managers accountable for their investment decisions are in any case difficult. 

Illegal embezzlement and bribery, and vote-buying are the traditional forms of corruption in developing countries. Undue influence and influence buying through lobbying to distort political outcomes. Such purchase of influence is often legitimate and its activities are perfectly legal. One example is revolving door. Even businesses (like say the tech firms) may not even view influence peddling actions to change the rules of the game as corruption. 

Take the case of lobbying. It becomes corruption only when it crosses the line to advance narrow interests at the expense of society. But how we do know this is happening, especially when cause and effect are difficult or even impossible to disentangle? When does lobbying cross the line to become corruption? 

How do we know that there was no corruption when Lloyd Blankfein and Co had such privileged access to Hank Paulson when he was designing a program to bail them all out? How do we know that there was no corruption when Obama met the Chief Executives of Honeywell and GE 30 and 22 times respectively in the 2009-15 period and those visits were accompanied by favourable regulatory or other benefits and abnormal share price increases for the respective companies? 

Bribery is easy to identify as corruption. But less easily identifiable forms of corruption, which are likely more pervasive in developed countries, may actually be much bigger and more corrosive.