Do derivatives make the financial system more efficient?

Sometimes it’s taken as a given that there wouldn’t be two parties trading derivatives, if the trade didn’t make both parties better off.   Because the existence of the trade counts as the evidence of efficiency, it’s not uncommon to see conclusions like the following:

[D]erivatives are used by financial institutions and corporations to adjust their exposure to particular financial risks, such as the default of a borrower or wild swings in interest rates. Both in theory and practice, these products have made the hedging and exchange of risk in the financial system more efficient.

It is clear that one of the implicit assumptions in the theory underlying the claims that derivatives make the financial system more efficient is that the parties to the trade have similar information. It is also abundantly clear that in practice this is not always true – as documented in recent lawsuits derivative markets are rife with asymmetric information.   Another issue is externalities – it is likely that the systemic effects of trade in derivatives are not taken into account by the counterparties to any given trade. For these reasons I think the question of whether derivatives make the financial system more efficient merits more careful analysis.

In my view, while there are many circumstances in which derivatives enhance efficiency, one cannot make a general statement that derivatives contribute positively to economic efficiency. In particular, our current derivative markets are very successful at transferring credit, interest rate and even liquidity risk (via collateral requirements) out of the financial system and onto the real economy. The question that must be asked is whether we want the real economy to bear these risks or whether we are better off forcing the financial system to bear these risks.

I’m going to propose a theory:  The role of financial intermediaries is to act as the economy’s managers of the risks involved in lending.

Traditionally there are two principal ways that intermediaries manage these risks: the first is to make a loan and carry all the risks of the loan on the balance sheet until the loan is paid off, the second is to sponsor a borrowing firm in its efforts to raise money on stock or bond markets. The first kind of intermediary is called a commercial bank, and because the bank carries the loans on its balance sheet until maturity, the traditional commercial bank can generally be counted on to manage the risks of lending carefully. The second kind of intermediary is an investment bank; traditionally an investment bank’s reputation for carefully vetting the quality of the stock and bond issues that the bank chose to underwrite was crucial to its ability to place the issue. Thus, by first approving the stock or bond issue and then by making a market in the stock or bond after it was issued, the investment bank played a very important role in managing the risk of the security.

We learned in the 1920s the dangers of market-based lending: because a stock (or bond) issue almost always relies on the imprimatur of a trusted bank or other gatekeeper, that gatekeeper can make a fortune by abusing the public’s trust. In many ways this is the same lesson we have relearned in recent years. While market-based lending has the advantage that it enables firms to raise very large amounts of money from a broad base of investors, this broad base of investors also generates the greatest weakness in market-based lending: it depends on gatekeepers who may or may not do their jobs well.

Thus, from the point of view of the economy as a whole the safest form of lending is lending that sits on the balance sheet of a bank until maturity. This lending is safe precisely because the bank is exposed to all of its risks, and therefore can be expected to manage them. There are a couple of provisos to this last statement. First of all, liquidity risk has always been something that individual banks can not manage entirely on their own. Reasonably stable banking systems have either developed means to deal with liquidity risk cooperatively or they have government sponsored support to help them through liquidity crises. Secondly, there is little historical or contemporary evidence that banks which have to manage interest rate risk without government support are willing to make long-term fixed rate loans. On the other hand, a very long history of banking and banking systems demonstrates that banks can manage credit risk and interest rate risk of three to five years very effectively.

Given this understanding that banking systems serve the economy by carrying the risks of making loans on their books – and indeed their main form of compensation, the net interest margin, is derived from the fact that banks carry that risk – we can now ask: Do derivatives make the financial system more efficient?

Let’s separate this into two different questions: Is the economy well served by credit default swaps that allow the financial system to transfer credit risk away from the banks? Is the economy well served by interest rate swaps that allow the financial system to transfer the risk of interest rates changes outside the banks?

Just as we learned both in the 1920s and more recently that market based lending is less reliable than bank based lending precisely because the bank is transferring the lending risk to other parties, we are likely to find that derivatives create a less reliable financial system than would be the case without derivatives precisely because derivatives make it easy for banks to transfer the risks of lending to others. Like the gatekeepers of market based lending, banks that transfer their risks using derivatives may at times do so in ways that abuse the trust of their counterparties.

In short if carrying credit risk on their balance sheets is one of the most important roles that financial intermediaries play in an economy, then making it easy for banks to transfer credit risk outside the banking system is likely to be destabilizing rather stabilizing for the simple reason that credit risk will end up being carried by entities that don’t know how to evaluate and manage it. This would be one reason to doubt that credit derivatives make the financial system more efficient.

Interest rate swaps are more complicated, because the US now has a financial system where it is considered normal for banks to make 30 year fixed rate loans. This system was created in the 1930s as a way to deal with a vicious cycle of homeowner insolvency: because so many homeowners couldn’t refinance their five year interest only mortgages when the principal payment was due, homes had to be sold and there were so many sellers that house prices were seriously affected – guaranteeing that even more homeowners couldn’t refinance. The solution to this vicious cycle was the thirty year fixed rate mortgage – except for the fact that there was no such product on the market.

I’m not sure when precisely the thirty year fixed rate mortgage became common, but the 1930s saw vast government efforts to support the residential mortgage market. In 1932 Herbert Hoover signed into law the Federal Home Loan Bank Act, which established Federal Home Loan Bank System, a mutual organization that raises money on public markets and stands ready to buy mortgages from member savings and loan banks. (It was modeled on the Federal Reserve System which through its discount window stands ready to buy commercial obligations from member banks.) In 1934 the Federal Housing Administration (FHA) was established. The FHA offered government mortgage insurance on loans that met the standards of the FHA. Because banks were reluctant to buy FHA mortgages, in 1938 the Federal National Mortgage Association (Fannie Mae) was created as a government agency that bought FHA mortgages. Somewhere along the line the thirty-year fixed rate mortgage became a staple of the industry. In the absence of substantial government sponsorship of this mortgage product it’s hard to imagine that banks would be willing to offer such mortgages or hold them on their balance sheets.

The savings and loan crisis of the 80s was caused by the fact that these banks were carrying 30 year fixed rate mortgages on their books and were completely unprepared to deal with the interest risk created by their long duration assets. (No bank can make money if it has to pay 10% on deposits and only earns 5% on its portfolio of loans.) The interest rate swap market – which was pretty clearly the catalyst for the more general growth of modern derivatives markets – developed as a response to the savings and loan crisis and the recognition by financial institutions that their long duration assets exposed them to interest rate risk. Interest rate swaps allow banks to convert their fixed rate assets into floating rate assets and thus have an important role to play in US markets as they are currently structured.

The savings and loan crisis can, however, be viewed as a sign that from a systemic point of view that 30 year fixed rate mortgages are simply too dangerous a product. Shifting towards a shorter duration mortgage product or one with some measure of floating rates (and there are many, many examples of such products abroad) might have been a better choice for the financial system. Thus the problem with interest rate swaps may be that they encourage individual banks to “manage” interest rate risk, when in fact it is far from clear that the financial system as a whole is equipped to deal with the aggregate exposure to interest rate risk that the 30 year fixed rate mortgage creates.

In short, the problem with derivatives is precisely that they allow financial institutions to transfer risk. In the case of credit derivatives the problem is that credit risk is too easily transferred away from the financial institutions that are specialists in credit risk to non-specialists who don’t really understand the risks they are carrying. In the case of interest rate swaps, the problem is that the perception that interest rate risk can be traded away leads to the growth of excessive interest rate risk in the financial system as a whole. It is not unlikely that when there is actually a significant jump in interest rates, many of the counterparties that are carrying that risk will be bankrupted by it and we will find that once again the government needs to step in as the derivative underwriter of last resort in order to prevent a chain of financial system failures.   (In fact, as I think about the case of AIG, this argument about the externalities of interest rate swaps seems to hold pretty well for credit derivatives too.)

The captive minds of finance

Note (7-26-10): In this post I’m just using Adair Turner’s Project Syndicate piece as a means to critique the argument that financial innovation “completes markets” and therefore contributes to efficiency of our financial markets.  Lord Turner is definitely not one of the “captive minds” of finance.  He’s one of the people who’s working to reform the system.  However, I think the fact that he’s doing so from within the system means that he’s sometimes a little too soft spoken.  I’m just trying to point out how utterly ridiculous the ideology underlying deregulation really was.

In a critique of the economic ideology of the past era, Adair Turner writes:

Indeed, at least in the arena of financial economics, a vulgar version of equilibrium theory rose to dominance in the years before the financial crisis, portraying market completion as the cure to all problems, and mathematical sophistication decoupled from philosophical understanding as the key to effective risk management. Institutions such as the International Monetary Fund, in its Global Financial Stability Reviews (GFSR), set out a confident story of a self-equilibrating system.

Thus, only 18 months before the crisis erupted, the April 2006 GFSR approvingly recorded “a growing recognition that the dispersion of credit risks to a broader and more diverse group of investors… has helped make the banking and wider financial system more resilient. The improved resilience may be seen in fewer bank failures and more consistent credit provision.” Market completion, in other words, was the key to a safer system. …

[A]t regulatory agencies like Britain’s Financial Services Authority (which I lead), the belief that financial innovation and increased market liquidity were valuable because they complete markets and improve price discovery was not just accepted; it was part of the institutional DNA.

This talk of market completion makes it sound as if there is some formal economic theory that supports the view that incremental “market completion [is] the key to a safer system”.  The point needs to be made that only people who don’t actually understand economic theory could possibly make this argument.  Unfortunately if Adair Turner’s description of the culture at the FSA is correct, then the problem that he is describing is not one of the dominance of a specific school of economic thought, but instead the complete failure to use the tools of economic analysis.

The concept of market completeness comes from the Arrow-Debreu model, which is itself the modern apotheosis of classical economic theory as it was first posited by Adam Smith and then developed formally by Leon Walras.  In the Arrow-Debreu model every member of the economy can buy insurance against any possible eventuality in that economy.  There is no such thing as bankruptcy, but only entities that (i) understand what they will and will not own in every possible future eventuality and (ii) make perfectly credible promises to share what wealth they have in “tail risk” settings as long as they are paid a fair premium to do so beforehand.  Anybody who has studied this model understands that “complete markets” refers to an idealized world that does not, and indeed, cannot ever exist.

The Arrow-Debreu model was developed more than half a century ago, so economists have had plenty of time to study the question of whether in an economy with incomplete markets (that is, in an economy with some missing insurance markets) the addition of another insurance contract is necessarily welfare improving.  The answer to this theoretical question is incontrovertibly no (as Lord Turner acknowledges in a middle paragraph).  While there may be some special circumstances where increasing market completeness may enhance economic efficiency, in the general case, there is no reason to expect an increase in economic efficiency.

Furthermore, even before the work on incomplete markets was written, it would have been an obvious logical fallacy to jump from the observation that an economy with complete markets is efficient to the assertion that as markets become more complete, they become more efficient.  Thus the failure of the culture at the IMF and the FSA that is described by Adair Turner is a failure to apply basic economic theory to the problems they faced.  And Adair Turner is far too kind to both members of the financial industry and to his colleagues at the FSA:

Market efficiency and market completion theories can help reassure major financial institutions’ top executives that they must in some subtle way be doing God’s work, even when it looks at first sight as if some of their trading is simply speculation. Regulators need to hire industry experts to regulate effectively; but industry experts are almost bound to share the industry’s implicit assumptions. Understanding these social and cultural processes could itself be an important focus of new research.

But we should not underplay the importance of ideology. Sophisticated human institutions – such as those that form the policymaking and regulatory system – are impossible to manage without a set of ideas that are sufficiently complex and internally consistent to be intellectually credible, but simple enough to provide a workable basis for day-to-day decision-making.

If financiers and regulators relied on theories of “market efficiency and market completeness” to support their activities, then that is prima facie evidence that they did not understand these theories.  There was nothing “internally consistent” or “intellectually credible” about the claim that the assets the financiers were creating and the regulators were failing to regulate were efficiency enhancing for the economy as whole.

I can certainly agree with Adair Turner that social and cultural processes that drive people to share an ideology and make common assumptions have played a huge role in this crisis, but I think it is a mistake to dignify the process of rationalization that took place as “internally consistent” or “intellectually credible”.  There was an ideology and there was a pseudo-logical foundation to that ideology; these foundations crack easily, however, as soon as they are subject to careful analysis.  It is important to recognize that while ideology likely played a large role in the crisis, logical coherence did not.

Do budget surpluses cause private sector indebtedness?

Marshall Auerback (h/t Naked Capitalism) doesn’t quite say, but strongly implies, that the Clinton era budget surplus caused the rise in private sector indebtedness that has caused so many problems.

I think there’s a sense in which this is true, but  that the underlying problem is excessively high growth targets (i.e. expectations of economic performance).  The politicians wanted declining budget deficits and strong growth in an economy where the banking system was already undercapitalized.  In order to square the circle, the regulators changed the rules so that more debt could be issued with lower levels of capital — that is, the regulators approved the growth of the shadow banking system and the use of “hybrid capital” (=debt masquerading as equity) — because there was no other way to square the circle.

And here we are.  I just hope it’s possible to learn from history.

Assume a can-opener …

Chris Edley (via Mark Thoma) writes that Treasury should advance funds to the states:

Of course, when Treasury eventually collected what it was owed, the state would have to cut spending or find new revenue sources. But that would happen after the recession, when both tasks would likely prove easier economically and politically.

Arguments that are premised on the inevitable return of economic growth to rescue us from our folly sound remarkably similar to what was being said in the 20s and early 30s.  (Revisions of German reparation plans and the CreditAnstalt’s bailout of the Bodencreditanstalt bank were expected to work because of the coming economic recovery.)

Given the state of the world economy right now doesn’t it makes sense to work on a solution that can have a moderate degree of success even if we have economic stagnation for a decade or two.

Does anyone really believe that this is a liquidity crisis?

Xavier Gabaix argues that restrained business investment is a natural — and healthy and problematic — reaction to macroeconomic tail risk.  And Brad DeLong responds:

The macroeconomic tail risk that businesses today fear is not a Great Forgetting of technology and organisation or a Great Vacation on the part of the North Atlantic labour force … [These] are the only two “macroeconomic tail risk” shocks I can think of that would make it socially and collectively rational for businesses as a group to refuse to invest in new capacity right now

DeLong’s references make it clear that he persists in viewing our current problems as a liquidity, rather than an solvency crisis.  [Say explains that the Bank of England’s tight monetary policy in 1825 triggered a sharp recession by curtailing the supply of credit firms and causing them to hoard cash.  Mill explains that it was only after the Bank reversed its policy and loosened the money supply that the economy recovered.  And the Kindleberger reference makes the liquidity interpretation entirely clear because a “lender of last resort” only functions because it lends to solvent firms — a condition that Mill also emphasizes.  For more on the crisis of 1825 see here.]

In short, Prof DeLong appears to imagine that if we continue to treat the crisis of 2007 as a liquidity rather than a solvency crisis — eventually the solvency crisis will disappear.  The theoretic foundations of this view are hard to fathom — especially given that our financial institutions have been relying heavily on “liquidity support” for almost three full years — could there be any stronger evidence that we are facing a solvency crisis, not a “panic”?

[The timeline of the 1825 crisis is as follows:  March 1825 Bank of England starts to tighten, December 1825 full blown bank crisis with money center bank failures, met with aggressive lender of last resort action, by June 1826 lender of last resort activities have wound down and by December 1826 monetary measures and bankruptcies have normalized.  See Chart 11 here and Chart 1 here.]

In my view business investment is restrained not because of macroeconomic tail risk, but based on elementary macroeconomic analysis:  As long as policymakers are determined to prevent the bankruptcy of insolvent firms, necessary economic adjustments will be delayed and valuable assets will be entombed in firms without the balance sheet capacity to borrow the funds necessary to realize the value of those assets.  This policy will have the effect of preventing the economy from growing at its natural rate over the next decade and render investment a very dubious prospect indeed.

In short, what’s holding corporate investment back may be uncertainty over future policy and the nature of the institutional structure of the economy.  That is, firms may fear that bad economic policy in the form of using liquidity tools to support insolvent firms until they become solvent — in the next decade or two if ever — (similar to the Japanese “solution” of avoiding the restructuring of banks and firms) is going to cost the economy so much that it really doesn’t make sense to invest in a future without economic growth.

Update 7-7-10: Gabaix points out that it is possible that the future policy that firms are worried about may be a failure to bailout in the next crisis.  Is this the same point I am making or a different one?  Hmm.

And Krugman argues that there’s no need to interject “future government policies” into the discussion at all — a review of data suggests that investment is higher than one would predict in an environment where we simply don’t need that much new construction, etc.

I am happy to acknowledge that I have no way of determining the true source of concerns about the future economic performance, but I hope everyone can agree that too much government intervention, too little government intervention and misdirected government intervention are all on the list of  possibilities.  For which reason the decision-makers who are actually navigating these shoals have all my sympathy.

What drives low corporate investment?

The Economist asks “What explains the rise in corporate thrift?”  I think the question needs to be more carefully defined.  To what degree is the “rise in corporate thrift” driven by the hoarding of cash and to what degree by an effort to reduce the overwhelming burden of debt?  When firms are observed “hoarding cash”, is this a reflection of a movement of their financial investment portfolios into cash or a genuine decision to reduce internal investment?  Sharpening the question via a more careful breakdown of firm behavior would improve the quality of the discussion.

Another factor that needs to be considered is whether we are talking about the decline in corporate investment over the last decade or so or the more recent jump in corporate savings.  While the Economist is clearly addressing the recent rise in corporate savings due to the crisis, Yves Smith and Rob Parenteau have an op-ed in the NYT that looks at longer term trends in corporate investment.  Their thesis that “incentives for both managers and investors now favor myopia and speculation, undermining the very operation of capitalism” merits further analysis.