From “Analysis of the 2008-2009 Financial Crisis”, by Terry Marsh and Paul Pfleiderer:
In this Preface, we offer some analysis of the 2008-2009 financial crisis and its implications for financial industry reform and research. We primarily focus on issues relating to transparency and the measurement of risk and how these are affected by management incentives that are often misaligned with the incentives of those who are exposed in various ways to the risk being measured. In the aftermath of the crisis many have called for increased transparency; we suggest that while transparency is no doubt a desirable goal in many ways, enhancing it could prove to be quite difficult.
I found this particular paragraph of interest:
As established economics tells us, and as we are reminded almost daily by the financial media, incentives to take on risk are magnified in “heads I win and tails someone else loses” situations. In particular, implicit government insurance (e.g., too-big-to-fail insurance) potentially leads financial institutions to take on socially undesirable levels of risk, unless the government can somehow magically
and instantly adjust for the level of risk taken and perfectly price and charge upfront for the implicit insurance it provides. Financial liberalization permitted the financial sector to take on more leverage and more risk, and the incentives of managers and others all but guaranteed that would happen. There is certainly supporting anecdotal evidence of “risk-be-damned” behavior by large financial players prior to the crisis, e.g.: “[A former senior Lehman banker] recalls vividly the days in early 2007 at Lehman when his financial models began to throw up more warnings showing delinquencies and defaults, and he remembers colleagues on his desk raising questions about loan quality. But he said the firm’s ranking as
the top loan originator on Wall Street, not to mention the pressures put on the desk by Lehman’s growth-obsessed leadership, made it difficult for even the most senior executives to raise questions” (New York Times, September 13, 2008 p. 7).Of course, different financial institutions have different cultures and compensation plans. All that economics can say is that implicit government insurance,
among other things, creates incentives that can lead to more-than-otherwise prudent risk-taking. Management hubris or over-confidence undoubtedly played a role — the bail-out insurance simply lowers the downside costs of that behavior. The stunning magnitude of the bail-outs that did occur make it plausible that they played a major role in prior decision-making: Alessandri and Haldane (2009)
calculate that “the scale of intervention to support the banks in the UK, US, and the euro-area during the current crisis…totals over $14 trillion or almost a quarter of global GDP. It dwarfs any previous state
support of the banking system” (p. 2).
After reading many of the recent accounts of the lead-up to the financial crisis, I do wonder exactly how much of the problem was with the models per se, and how much was organizational/institutional incentives to ignore unwelcome news brought to executives by the modelers.
In any case, a paper well worth reading, adding to a crisis reading list.
a question to ask: if i, as a modeler, was skeptical about the models used or had models that showed materially different outcomes, could i have maintained employment?
i’m sure you can guess the answer.
so the answer is that, yes, modelers were just there to say yes, and institutional incentives were primarily to blame.
I wonder why “established economics” doesn’t tell us more explicitly that the magnitude of manager compensation is incompatible with financial stability.
“Heads I win, tails you lose” is the game financial managers and executives are playing with stockholders, customers AND the government, when, year in and year out, they bank in salary and bonuses money that ought to be be adding to the capital cushion hedging the risks they are supposedly insuring. Instead, in order to pay themselves billions and billions, they have — as most notoriously in the AIG case — sold insurance without actually accumulating the reserves to back it up. Instead of accumulating such reserves, executives and managers took the money and ran.
Rather than focusing so much on the undoubtedly important question of gov’t insurance, it might be wise to try to answer the fundamental question of just how much finance an economy needs. How much of national income needs to funnel through the financial sector? What is the economic function of the financial sector, and are there any measurements of how well it accomplishes those functions?
Because through the fog of ungrounded abstract assertions about risk and return, I am not seeing a whole lot of upside from the growth of the financial sector over the last generation. Are pensions more secure? University endowments? Health insurance getting cheaper and more available? Credit card interest rates driven to rock bottom by ruthless competition? Venture capital driving business formation around marvelous innovations?
Not really feelin’ it? Me, neither. Let’s look for some well-grounded perspective, that does not begin and end with some right-wing crapola meme, about the noxious effects of government-provided insurance?
The early days of 2007? Those were hardly early. The subprime crisis began in earnest in mid-January, when HSBC and New Century Mortgage issued back-to-back warnings on rising subprime delinquencies. In fact, I those delinquencies were rising steeply through the second half of 2006, a trend which New Century referred to as the mere “normalization” of delinquencies from the ultra-low levels of 2003-2006. So if “the modelers” were “raising questions about loan quality” in “the early days of 2007”, they could have done so just by reading the Wall Street Journal. Where were “the modelers” when those subprime loans were being made? They were arguing that their regressions showed income documentation was not correlated with defaults, which is why no-doc (“liar”) loans grew to 65% of all subprime loans in 2006. This was truly a failure of at least a subset of econometricians.
The Financial Time’s Gillian Tett (who has a Ph.D, in Social Anthropology from Cambridge) in her recent book,
Fool’s Gold, provides another example of the cultural aspect. She writes about how JP Morgan Chase after doing the analysis, decided not to get into mortgage linked CDOs and the like. And this was after knowing that it was costing the JP Morgan Chase over a billion dollars in lost revenue. In other words, the culture of the organization was to take risk seriously.
Read “The Sellout” by Charles Gasparino if you still have any doubts about whether it was models or compensation that ruled Wall Street firms.
The book is not very scientific, just a reporter’s account of the industry over the last 30 years. He tracks the careers of the now famous top execs on Wall Street. (Cayne, Fuld, Fink, Vranos, Meriwether, Milken, Mozilo, Ranieri, Rubin, Spector, Thain, Weill, and all the rest.)
Amazing how similar 1987, 1994, 1998, and today are. They just got bigger each time. And securitization, combined with massive leverage, justified by models to obfuscate risk and re-sell it to the rest of the financial world was the core problem each time.
The author made a good point about models. Meriwether, founder of LTCM, was regarded as the wiz of risk models. He considered his models “safe” 99% of the time. The author pointed out that the chances of an airplane crash are 50 million to 1. But would you fly if the chances of a crash are “only” 100 to 1? Russia defaulted, TLCM went insolvent, and that was the grandaddy of “systematic risk”. The hedges didn’t work. A combination of a Wall Street firm led bailout (since they would go under from “contagion”) and the beginning of the Greenspan “put”, made moral hazard a reality from then on.
This go around, the only one left to do the bailouts was the government.
Unwelcome news from modelers is easy for managements to ignore when there is no downside risk from ignoring it. Banish TBTF and the Bernanke put.
In particular, implicit government insurance (e.g., too-big-to-fail insurance) potentially leads financial institutions to take on socially undesirable levels of risk, unless the government can somehow magically and instantly adjust for the level of risk taken and perfectly price and charge upfront for the implicit insurance it provides. Financial liberalization permitted the financial sector to take on more leverage and more risk, and the incentives of managers and others all but guaranteed that would happen.
Menzie,
I agree that this is a very interesting paragraph. Thanks.
My only comment is that the implication is that the risk taken was an “unintended consequence” of governmental action. It seems pretty clear that the risk taking was an “intended consequence” of the government. They forced lending institutions to loan to bad risks and then took the risk off of the primary lenders through Fannie and Freddie and other programs, not to mention the bailouts that took on the bad risks the institutions did not have time to pass on to the government.
The other big heads I win, tails you lose incentive that encouraged risk was the widespread use of stock options for compensation, which have that character. Stock options, of course, are strongly favored over other alternatives by the tax code.
RicardoZ,
Why do you say that the government forced banks to lend to sub prime borrowers?
Other than the fact that secure investments didn’t pay as much interest as bad risk borrowers, I don’t see any “forcing”.
Perhaps I’m missing your point?
Indeed, although in an inflationary environment there is a moral hazrd for any, even the honest, debtor: to alleviate himself from his debt through the effects of currency debasement. Whats wrong with Economics?