Tuesday 19 August 2008

It’s a wind-up

As the credit crisis has unfolded, several vicious circle processes have come to light which have contributed to its magnitude and persistence. Such positive feedback loops – which in this case are sometimes wrongly called negative feedback loops because they reinforce the credit crisis – act to increase the response of a system to a change in an input variable. On the assumption that these mechanisms are intolerable, analysts of the credit crisis have suggested various financial system reforms to eradicate them. RebelEconomist, however, questions whether such positive feedback loops really are a major problem, and considers that some of the proposals to get rid of them might create their own perverse effects.

One such positive feedback mechanism that has been much discussed recently is bank capital adequacy regulation. The existing fixed capital ratio rule has the effect of amplifying the business cycle, because during a period of elevated credit losses like the present, banks may prefer to contract lending and sell existing assets to reduce any resulting capital shortfall rather than sell additional shares at depressed prices or cut dividends to raise more capital. To the extent that selling and buying less assets reduces asset values in general, the asset side of the banks’ balance sheets contracts further, and forces further asset sales, and so on in a vicious circle. In order to reduce this inherent pro-cyclicality, Goodhart and Persaud propose that a bank’s capital requirement should depend on its recent asset growth, so that bank capital coverage automatically rises during credit booms to restrain the expansion, and falls in busts. The trouble is, who would lend money to a bank that was being allowed to run down its capital cushion against losses at a time of financial turmoil? You might as well propose that gravity be reduced at the bottom of hills. One lesson that should have been learned from the credit crisis is that, if there is the slightest doubt about a bank’s solvency during a time of stress, it will be shunned by lenders. Unless the scheme introduces counter-cyclicality by forcing banks to hold excess capital at all times except the very worst, which would certainly be strongly resisted by the banks and may well be inefficient, it might even exacerbate a credit crisis.

Another positive feedback mechanism is the requirement for financial businesses to report realistic market, or “fair”, values for a large proportion of the assets they hold (the alternative being to value the assets at their historic cost when acquired). This requirement (eg as specified in FASB statements 157 and 159 in the US) has been applied increasingly extensively to financial businesses. Since fair market valuation reveals the effect of falling asset values on firms’ solvency, it raises their cost of funding following such losses and so exacerbates their problems. In this case, proposed remedies include restricting the scope of fair value accounting to assets that are intended to be held for a short period only, or at least suspending further introduction of fair value accounting until the credit crisis subsides. No doubt a firm in difficulty would be grateful for the opportunity to try to trade its way out of trouble with money owed to its creditors, under the cover of accounts that do not reflect creditors’ exposure to a potential default, but such obfuscation could be expected to discourage future commitments and raise the cost of new credit. In fact, economic theory suggests that, in the absence of full disclosure, creditors might well assume that the danger is worse that it really is. Apart from providing some short-term relief at the expense of already committed creditors, abandoning fair value accounts might also exacerbate a credit crisis.

Perhaps the concern about positive feedback mechanisms in the financial system is misplaced. Positive feedback is quite common in nature, such as the amplifying effect of sea ice on global temperature or the cascade of neutrons in nuclear fission. No doubt, as discussed, many of the positive feedback mechanisms in the financial system exist for good reason. Provided that a feedback process is convergent, or is interrupted or counteracted by some other process before it progresses too far, it is not necessarily a problem. It can simply be seen as playing a role in determining the path and amplitude of the system’s dynamic response to changes in input variables. What really matters is that the influence of any such feedback mechanisms is taken into account when predicting the response of the system to shocks. In short, a reasonable answer to coping with financial system positive feedback mechanisms like capital adequacy requirements or fair value accounting is for firms to allow for them by either further reducing their exposure to risk or building in more protection.

11 comments:

Anonymous said...

"...restricting the scope of fair value accounting to assets that are intended to be held for a short period only"

i.e. trade; if it's profitable, keep it in the trading account, and if not, stuff it in the investment account so it's not MTM under the relevant FASB regs. This is akin to the taxpayer giving free options (on contingent capital usage) to banks.

Anonymous said...

I have repeatedly read that a lesson of the Japanese banking/credit crisis was the importance of prompt disclosure and write-offs. This would seem be in direct contradiction to the suggestion of suspending MTM accounting.

RebelEconomist said...

Anonymous #2,

I totally agree. I think that the US has learned totally the wrong lesson from Japan's post-bubble experience. By trying to brake the correction before it was complete, Japan used up most of its stimulation options before the trough was reached, so that when they did seem to hit the bottom after about fifteen years, the Japanese authorities had nothing left to decisively get their economy going again.

I wonder if the right answer is to begin with Andrew Mellon's infamous depression advice - ie "liquidate, liquidate" - followed by very aggressive fiscal and monetary easing. While the result might be, say, a 20% decline in GDP in a year or two, followed by a rapid recovery, this might well be preferable to persistent weakness, with output growth a percent or two below potential for years, as in Japan.

Anonymous said...

Rebeleconomist
Hopped over from the pettis site

btw, just a thought but could one of the ways of building up protection for fair value accounting be more "opaque" accounts?

btw, am old enough to remember the existence of east germany and the way the west was dragged kicking and screaming into a reunion. do you really think modern germany could have seen the growth of the last few years without it?

RebelEconomist said...

Judy,

I am a little confused by your first point......who is building up protection and against what? Whatever, I have yet to hear a good argument for opaque accounts, except perhaps that fair value accounts can be more costly to produce than using historic cost. Even then, I would say that any competent manager would want to have at least some idea of what the present market value of their assets is worth, if only to review their past decisions, and any potential creditor would want some idea of the net worth of the firm before they committed themselves, so the information should be produced anyway.

I am not an expert on Germany, but it seems to me that the underlying reason for Germany's recent success is that they produce quality goods that customers all over the world are willing to pay up for. That and the fact that Germans avoid the kind of complacency that led to debt-financed housing bubbles in the US and UK. Germany's Agenda 2010 reforms got their public finances back under control, and wage restraint restored German competitiveness in Europe despite a fixed exchange rate (a lesson for the US in its economic relationship with China?). In short, the Germans know how to chi ku! Since similar economic discipline helped West Germany to come through the seventies relatively successfully, I do not think that reunification is primarily responsible for Germany's recent success, although of course the former East German Lander will have been able to grow relatively fast by catch up.

Anonymous said...

What do you think of the recent capital insurance proposal presented at Jackson Hole?

RebelEconomist said...

Anonymous at 01 September 2008 08:49,

I presume that you mean the paper by Kashyap, Rajan and Stein (KR&S)? They make a good argument about the potential for bank capital to be abused when it is not needed, so their insurance idea sounds like a useful contribution to me. I can see a couple of problems with it though. First, as KR&S admit, like any insurance, it may generate moral hazard. For that reason, it seems to me that the insurers would want some right to inspect and impose conditions on banks that may duplicate regulatory banking supervision. Second, I would be cautious about securitising such insurance. As KR&S say, the market has a demand for securities with payoffs consisting of a usual small positive return balanced by the occasional large loss. I suspect that this is because such payoffs are attractive to agent fund managers who are paid a cut of positive returns but can only be fired in the event that the large loss is realised. Instead of having to bail out the banks during a financial crisis, the government might end up having to rescue pension funds and insurance companies instead.

Anonymous said...

"do you really think modern germany could have seen the growth of the last few years without it?"

Yes. German growth has been driven by exports (to EM) and investment (probably a function of export demand). I don't think reunification has anything to do with either of those. To that I would add that, as RebelEconomist pointed out, Germany has become much more competitive as a result of a long-period of declining real wages leading to higher productivity growth. One might argue that has something to do with reunification.

Anonymous said...

Re: the Jackson Hole paper --- I don't see what the fuss is about. Myron Scholes proposed what is essentially the same idea 10 years ago (I mentioned this in the comments section on Dave Altig's MacroBlog). That nobody ran with it should tell you something.

Anyway, if you're interested, the reference is...Scholes, M., (1999), “Liquidity options: Scholes explains”, Risk, Volume 12, Number 11, November, pp. 6

RebelEconomist said...

Anon.,

Thanks for the reference. I did see your comment (as MW) on macroblog, and appreciated your following through with the reference. I wish all comments on economic blogs were as thorough.

Anonymous said...

Re: contingent capital options --- I think the issues are 1) the option is probably unhedgable and 2) it would be exercised at the worst possible time for the writer. Scholes suggested pension funds (PFs) as natural writers. But consider the current environment of falling equity prices and rates, which is the worst of all worlds for PFs. They clearly wouldn't want to have to pay out on a contingent capital option right now (a problem perhaps exacerbated by regulation regarding surpluses/deficits being brought on to the balance sheet).

Anyway, there is a tangentially related discussion here: http://www.wilmott.com/messageview.cfm?catid=3&threadid=64530.