Posts this month
A blog on financial markets and their regulation
Companies that repeatedly pivot from one business to another (more glamorous) business could be indulging in a ponzi scheme designed to hide business failure and lead investors on a wild goose chase for an ever elusive pot of gold. There are some very large companies in India and in the United States about whom one could harbour such a suspicion.
The question is how can one distinguish these corporate ponzis from genuine pivots. After all it makes sense to change your business as situations change. Warren Buffet’s Berkshire Hathawy pivoted from the textiles business to insurance and finance and if its next elephant size deal is like its last one, it could pivot again to a non financial conglomerate. In India, Wipro became a software giant after a pivot from vegetable products.
One indicator of a ponzi is that the pivot typically chases a prevailing stock market fad rather than any particular competence or competitive advantage in the new business (unless one counts cheap capital as a competitive advantage). But even that is not determinative as the case of GE makes clear. As a Financial Times FT View pointed out a couple of months ago “In the dotcom bubble, GE was valued as a tech stock; in the credit bubble, it was valued like a leveraged debt vehicle (which, in large part, it was).” To which one could add that till recently it was trying to position itself as a leader in the industrial Internet of Things. That makes GE a stock market opportunist, but not a ponzi. Even after returning to its old industrial roots in the last few months, GE remains a valuable business.
The corporate ponzis that I worry about are something else altogether. This kind of company is a graveyard of serial failures, even though the future always looks rosy. In the heyday of each of these failed businesses, the market would not have bothered about current losses, because it would have valued the business on multiples of current or future revenues. After the company pivoted away from the business, the market would not bother about the losses (and revenue collapse) in the old business because the market is always “forward looking”. The corporate ponzi’s challenge is to find the next big thing (and make it bigger than the last big thing). When their luck runs out and the corporate ponzi finally fails, everybody wonders why nobody saw through the fraud earlier.
More than a decade ago, in the days before the Global Financial Crisis, I asked a provocative question on this blog: “Had we invented CDOs first, would we have ever found it necessary to invent banks?” (I followed up in the early days of the crisis with a detailed comparison of banks with CDOs).
I am revisiting all this because I just finished reading a fascinating paper by Juliane Begenau and Erik Stafford demonstrating that, banks simply do not have a competitive edge in anything that they do. Specifically, the return on assets of the US banking system over the period 1960-2016 was less than that of a matched maturity portfolio of US Treasury bonds. This is a truly damning finding because banks are supposed to earn a return from two sources: maturity transformation (higher yielding long term assets funded by cheaper short term financing) and credit risk premium (investing in higher return risky debt). What Begenau and Stafford found is that their actual return does not match what you can get from maturity transformation without taking any credit risk at all.
That raises the question as to why banks have survived for so long. Another finding of Begenau and Stafford can be used to provide an answer: maturity transformation (even without any credit risk) with typical banking sector leverage is not viable in a mark-to-market regime. The banking regulators have acquiesced in the idea that the loan book of the banks need not be subject to mark to market. Making illiquid loans and taking credit risk is the price that banks have to pay to become eligible for hold-to-maturity accounting of their loan book. Banks are able to undertake maturity transformation with high levels of leverage without wiping out their equity because the loan book is not marked to market.
Hold-to-maturity accounting allows banks (and only banks and similar institutions) to carry out leveraged maturity transformation. This competitive advantage means that banks are able to make money on maturity transformation. However, they are so bad in their credit activities that they lose money on this side of their business. This offsets some of the returns from maturity transformation, and so they underperform a matched maturity portfolio of risk free bonds.
It is important to keep in mind that credit risk earns a reliable risk premium in the bond markets. Therefore, if banks manage to earn a negative reward for bearing credit risk, it is clear that either their credit risk assessment must be very poor or their intermediation costs must be very high. Interestingly, Begenau and Stafford do find that maturity transformation using risk free bonds has no exposure to systematic risk (CAPM beta), banks have CAPM betas close to one. The credit activity of the banks creates risks and loses money; in short, banks are really bad at this business.
I have always been of the view that banks are an obsolete financial technology. They made sense decades ago when financial markets were not developed enough to perform credit intermediation. That is no longer the case today.
This is particularly relevant in India where we have spent half a century creating an over-banked economy and stifled financial markets in a futile attempt to make banking viable. The crisis of bad loans in the banking system today is a reminder that this strategy has reached a dead end. As I wrote nearly a year ago:
India needs to move away from a bank dominated financial system, and some degree of downsizing of the banking system is acceptable if it is accompanied by an offsetting growth of the bond markets and non bank finance.
After the recent multi-billion dollar fraud at a leading Indian public sector bank, there has been a chorus of calls in India for privatizing state owned banks. We would do better to shut some of them down. Time and money are better spent on developing a bond market unshackled from the imperatives of supporting a weak banking system.
Between early October 2016 (shortly before demonetization) and today, the Reserve Bank of India (RBI) has cut its policy rate twice (October 4, 2016 and August 2, 2017) to bring the repo rate down by 50 basis points from 6.5% to 6.0%. But the ten year Government of India bond yield is roughly 100 basis points higher than it was in early October 2016. Apparent monetary easing has been accompanied by a substantial tightening of financial conditions. This looks like a reverse of Greenspan’s Conundrum of 2005 in which the concern was that 150 basis points of rise in the US policy rate was accompanied by a falling trend in the long term yield.
Is it possible that the Indian situation could be a mild form of the bank-sovereign feedback loop?
The enhanced borrowing requirement of government causes a rise in government bond yields.
Rising bond yields cause more stress in the public sector banks because they hold a large amount of long term government bonds (unlike the private and foreign banks who tend to hold shorter term bonds). Rising bond yields may also act as a drag on the economy and worsen the non performing assets of the banks. In either case, the deterioration of the health of the public sector banks takes us back to Step 1 and the cycle can begin all over again.
If this analysis is correct, what can be done to break the bank-sovereign feedback loop? Several possibilities come to mind:
The RBI could take a leaf out of the Yield Curve Control policy of the Bank of Japan, and set monetary policy to prevent a rise in Indian government bond yields. In essence, the policy rate would no longer be the repo rate but the 10 year government bond yield.
The government could accomplish a massive pre-emptive recapitalization of the banking system that breaks the loop decisively.
The government could turn many of the public sector banks into narrow banks (or even shut them down) to eliminate the feedback loop.
The bank-sovereign feedback loop should not be a big problem for a currency issuing sovereign. This does not require any appeal to MMT, but is simply a reflection of the fact that banking sector liabilities are all nominal liabilities, and a currency issuing sovereign should not have any problem in backstopping these liabilities. If we still see evidence of such a loop, it should reflect some degree of mismatch between monetary policy, fiscal policy, and the bank recapitalization framework. And it should not be hard to fix the problem.
During the last week, the Indian financial sector has been gripped by the $1.8 billion fraud at Punjab National Bank (PNB). Fingers have been pointed at bank management, at the auditors and at the regulators, but finger pointing and angry denunciations do not solve problems. We did not solve the problem of unfriendly bank tellers by shouting at them; we solved it using technology (Paul Volcker once remarked that the most important financial innovation that he had seen was the ATM). That is probably the route we must take again: we cannot change human nature, but we can change the technology.
The blockchain technology that underpins cryptocurrencies like Bitcoin has the potential to reduce large banking frauds drastically because it enables radical transparency. Every transaction on Bitcoin is public and you do not even need a Bitcoin wallet to see these transactions. Many websites like https://blockexplorer.com/, https://blockchain.info/, https://www.blocktrail.com/BTC, https://btc.com/, and https://live.blockcypher.com/btc/ allow anybody with a web browser anywhere in the world to see every single transaction as it happens. We can use the same technology to allow the whole world to see every large financing or guarantee transaction (above some threshold like a billion rupees).
The shibboleth of bank secrecy can be discarded for large financing transactions because many of them become public anyway:
We could extend this into a uniform requirement to make large loans public:
The natural medium for such a disclosure is the blockchain. The alternative idea of using a credit registry has been an unmitigated disaster (just think of Equifax), and these agencies create more opaqueness than transparency.
If the PNB fraud pushes us to use the blockchain to make finance more transparent and therefore safer, $1.8 billion may end up being a price well worth paying.
Fabrizio Spargoli and Christian Upper have a BIS Working Paper with a different title: “Are Banks Opaque? Evidence from Insider Trading” with the following findings:
Our results do not support the conventional wisdom that banks are more opaque than other firms. Yes, purchases by bank insiders are followed by positive stock returns, indicating that banks are opaque. But banks are not special as we find the same effect for other firms. Where banks are special is when bad news arrive. We find that sales by bank insiders are not followed by negative stock returns. This suggests that bank insiders do not receive bad news earlier than outsiders. By contrast, insider sales at non-banks tend to be followed by a decline in stock prices.
My interpretation of the result is quite the opposite: banks are so opaque that even insiders cannot see through the opacity when bad things happen. Sometimes, as in the case of the London Whale, a market participant outside the bank has greater visibility to what is going on.
It appears to me that the findings of Spargoli and Upper are evidence that banks are too opaque to manage. Even a very competent chief executive can be clueless about some activities in a corner of the bank that have the potential to bring down the bank or at least cause severe losses. That would be an additional argument for moving from bank-dominated to market-dominated financial systems.
The following posts appeared on the sister blog (on Computing) during December 2017 and January 2018:
Why Intel investors should subscribe to the Linux Kernel Mailing List or at least LWN (Cross-posted on this blog as well)
Tweets during December 2017 and January 2018 (other than blog post tweets):
“Regulation is stealth taxation,” said US President Donald Trump at Davos yesterday. Can this taxation be Pigouvian, and can this stealth taxation be a good idea? That is the claim in Turk’s thought provoking paper “Securitization Reform after the Crisis: Regulation by Rulemaking or Regulation by Settlement?”
Turk argues that:
can been seen as imposing a Pigouvian tax on the specific market externality associated with securitization, and therefore come surprisingly close to a first-best policy intervention.
missed the mark because it was premised on a flawed theory of the role that securitization played the crisis, which emphasized traditional notions of fraud rather than poor risk-management.
It appears to me that there is no convincing evidence that securitization imposes large negative externalities requiring a Pigouvian tax. On the other hand, there is somewhat more evidence that banking creates large negative externalities, and Basel 3 is a kind of Pigouvian tax on banking. This Pigouvian taxation has also happened by stealth in the name of risk reduction.
We should worry about the knowledge deficit and the governance deficit in these exercises in stealth taxation. Regulators probably think that they have calibrated the Pigouvian tax correctly; but this is more likely to reflect conceit than genuine expertise in this field. Even if the expertise is granted for the sake of argument, the governance issue remains: can taxation be delegated to unelected regulators?
About a month ago, the US Federal Deposit Insurance Corporation (FDIC) published a 278 page document entitled “Crisis and Response: An FDIC History, 2008–2013.” It is a quite sanitized history compared to the excellent accounts of the crisis that came out many years ago (especially the books by Hank Paulson and Andrew Sorkin). Yet, I found that there was much of value in the FDIC book. There is of course a wealth of official and authoritative data, but there are also many interesting insights from the perspective of the regulators dealing with it in real time.
I wish Indian regulators could publish something similar about the various crises in Indian financial markets covering say 1990 to 2010 – the Harshad Mehta scam of 1992, the vanishing companies of 1995, the Ketan Parikh episode (especially the fate of the Calcutta Stock Exchange), the UTI Unit 64 bailout, Global Trust Bank, and Satyam. If the report of the Financial Crisis Inquiry Commission (FCIC) in the US did not affect the ability of the FDIC to publish their history, there is no reason why the reports of the Joint Parliamentary Committees (JPCs) should be an obstacle for the Indian authorities (RBI/SEBI/MOF/MCA) to publish their accounts of these episodes.
On January 3 and 4, 2018 (Wednesday and Thursday), the Intel stock price dropped by about 5% amidst massive trading volumes after The Register revealed a major security vulnerability in Intel chips on Tuesday evening (the Meltdown and Spectre bugs were officially disclosed shortly thereafter). But a bombshell had landed on the Linux Kernel on Saturday, and a careful reader would have been able to short the stock when the market opened on Tuesday (after the extended weekend). So, -1 for semi-strong form market efficiency.
Saturday’s post on LWN was very cryptic:
Linus has merged the kernel page-table isolation patch set into the mainline just ahead of the 4.15-rc6 release. This is a fundamental change that was added quite late in the development cycle; it seems a fair guess that 4.15 will have to go to -rc8, at least, before it’s ready for release.
The reason this was a bombshell is that
rc6 (release candidate 6) is very late in the release cycle where only minor bug fixes are usually made before release as version 4.15. As little as 10 days earlier, an article on LWN stated that Kernel Page-Table Isolation (KPTI) patch would be merged only into version 4.16 and even that was regarded as rushed. The article stated that many of the core kernel developers have clearly put a lot of time into this work and concluded that:
KPTI, in other words, has all the markings of a security patch being readied under pressure from a deadline.
If merging into 4.16 looked like racing against a deadline, pushing it into 4.15 clearly indicated an emergency. The public still did not know what the bug was that KPTI was guarding against, because security researchers follow a policy of responsible disclosure where public disclosure is delayed during an embargo period which gives time to the key developers (who are informed in advance) to patch their software. But, clearly the bug must be really scary for the core developers to merge the patch into the kernel in such a tearing hurry.
One more critical piece of information had landed on LWN two days before the bombshell. On December 27, a post described a small change that had been made in the KPTI patch:
AMD processors are not subject to the types of attacks that the kernel page table isolation feature protects against. The AMD microarchitecture does not allow memory references, including speculative references, that access higher privileged data when running in a lesser privileged mode when that access would result in a page fault.
Disable page table isolation by default on AMD processors by not setting the X86_BUG_CPU_INSECURE feature, which controls whether X86_FEATURE_PTI is set.
As Linus Torvalds put it a few days later: “not all CPU’s are crap.” Since it was already known that KPTI would degrade the performance of the processor by about 5%, the implication was clear: Intel chips would slow down by 5% relative to AMD after KPTI. In fact, one post on LWN on Monday evening (Note that Jan 2, 2018 0:00 UTC (Tue) would actually be late Monday evening in New York) did mention that trade idea:
Posted Jan 2, 2018 0:00 UTC (Tue) by Felix_the_Mac (guest, #32242)
In reply to: Kernel page-table isolation merged by GhePeU
Parent article: Kernel page-table isolation merged
I guess now would be a good time to buy AMD stock
The stock price chart shows that AMD did start rising on Tuesday, though the big volumes came only on Wednesday and Thursday. The interesting question is why was the smart money not reading the Linux Kernel Mailing List or at least LWN and getting ready for the short Intel, long AMD trade? Were they still recovering from the hangover of the New Year party?
Forbes India has an article on Bitcoin in the January 5, 2018 issue. It has the following quote from me:
Which is more crazy: That bitcoin has a market capitalisation of a couple of hundred billion dollars, or that 11 trillion dollars of bonds are trading at a negative yield, which means that people are lending money with the full knowledge that they will not even receive the full principal back let alone earn any interest? After the global financial crisis of 2008, many feel that the actions of central bankers have been reckless, and it is no wonder that these people are attracted to a currency that is not subject to the whims and fancies of central bankers. There is madness on both sides (fiat currencies of advanced countries and cryptocurrencies) and it is best to view both with equal detachment.
This is not the first time that I have stated the view that virtual currencies are a response to bad things happening in the real world (see for example, this blog post from October 2017).