Portrait of a Financial Crisis

Calculated Risk has a post this morning that really caught my eye and from which I derived the graphic above. You can click on the image for a larger version.

I’ve introduced a couple of trend lines. The first, in black, marks the trend for each bond classification from the beginning of the period in question to date and the second, in gray, marks the trend for each bond classification from the beginning of the period in question to the beginning of a serious deviation from trend and extends that to date.

First, look at the beginning of the obvious spike in yields. I measure that as Lehman Brothers filing for bankruptcy. Clearly, in hindsight although any number of people said so at the time, the Bush Administration made a serious error in its handling of the situation with Lehman Brothers. That spike isn’t just lines on a page. It’s a snapshot of a panic in the financial services sector. When you recognize that the spike indicates a sharp increase in the cost of borrowing for companies it also measures who knows how many thousands of jobs lost and who knows how many corporate bankruptcies that probably could have been avoided or, at the very least, slowed.

Over time the panic has subsided and we’ve largely returned to the trend.

But not entirely. Look at the difference between the spreads of the two bond classifications as they’ve actually materialized and as they would have been had they returned more or less completely to the prior trend.

Finally, note that much of the difference in the spread comes from Baa bonds paying a substantially higher yield.

All of these things suggest to me that, while the panic may be over, the graph above paints a portrait of a financial that is still ongoing.

I’d be remiss if I didn’t mention another alternative. It may be that the rating institutions are cooking the books and, as has been suggested, we shouldn’t pay any attention whatever to what they have to say.

However, after mulling over that op-ed a bit I’ve begun to wonder about it. When a guy who ran a bond rating agency tells you not to pay any attention to bond rating agencies, you’ve got to take pause.

I can’t help but wonder if he’s realized that the bond rating agencies have their own responsibilities for the problems that are ongoing. Is he doing damage control?

If bond rating agencies are selling snake oil, the solution isn’t to ignore them any more than the solution to dangerous and ineffective patent medicines wasn’t not to buy them (although that would certainly have solved the problem). The solution is to ensure that the rating agencies are doing their job. That’s something we might consider when we begin imposing new regulations on the financial sector.

Alternatively, we might insist they put up a bond of some sort to indemnify against the losses caused by wildly inaccurate ratings. And that brings us around to AIG. Who insures the insurers?

4 comments… add one
  • An instructive chart and a very thoughtful post, Dave. It provokes a number of thoughts– some directly responsive to your post, others a bit more tangential — so with your indulgence I think it’s time to earn my “emeritus” status in your blogroll. 😉

    I don’t think “trend” is a very useful way to think about the relative yields displayed in the chart. It’s not as if we would expect the trend lines to continue over the years — they don’t suggest to us where spreads are heading. The chart is just reflecting credit conditions and yield/risk relationships at various points in time. There is a “trend” to those relationships, but only in the sense that, at the start of the sequence, spreads were globally way, way, too skinny, reflecting a world awash in too much liquidity and an insane hunt for yield which compressed spreads among asset classes. As the chart show, over the past couple of years we’ve seen a correction of the extreme overshooting of credit prices.

    The chart depicts how we’ve been moving toward a new yield relationship between asset classes. It shows what is often referred to as a “re-rating” of asset classes, that is when there’s a major shift in risk/return expectations which the market has for a given asset class (Treasuries, investment grade corporates, high-yield, etc) relative to other asset classes. What we really want to know is when the re-rating process of corporates generally, and high-yield in particular, will be over — that is, when will your “trend” lines stop sloping upward.

    The chart suggests that risk is finally being (over?)priced and that liquidity preferences are so intense that, despite the amount of money being dropped from helicopters, we’re still extremely liquidity-tight. Your comments on the specific timing of when yields spiked and spreads gapped are on target. Your “trend” lines also highlight that after the Lehman shock, the system didn’t return to its pre-shock level. The new spreads reflected both a huge loss of confidence in the financial system — and therefore a willingness to pay big premiums for liquidity and low risk — as well as a post-Lehman recognition of a rapidly worsening condition of the real economy.

    Hopefully, however, spreads are starting to stabilize within a new liquidity/risk preference range, not continuing to gap wider, and any further deterioration of the corporate sector will be reflected in down-gradings of specific bond issues not in wider spreads among asset classes. We’re likely to see some bouncing around for awhile, however, reflecting occasional mini-panics that another financial sector shock may hit — e.g., nationalization of Citi that produces a big haircut for bondholders — until the US and European governments have convinced the markets there are no major surprises still buried in the banks or derivative markets.

    When the financial system eventually stabilizes, liquidity/risk premiums will start to decline and the spreads relative to Treasuries will narrow, reversing your “trend” lines (or perhaps better said, starting a new downward-sloping trend line). Though I expect it will be a few years before we see spreads over Treasuries get so skinny again, and I also expect the spreads among corporate asset classes will remain fairly wide for some time.

    As an aside, re the larger issue of what spread behaviors can tell us, as a rule of thumb, I use emerging market bond spreads as a rough indicator for when the global system is preparing for an implosion — when the spreads for marginal sovereigns (e.g. Central Asia) start heading sharply south (frex moving over a year from 800 to 400 bps), watch out. It means there’s a seriously excessive appetite for yield, buyers aren’t pricing risk, and somewhere there’s an unsustainable market that’s going to blow up and take a lot of seemingly uncorrelated markets with it. If I can see that, as at best a casual observer, it would seem to me that Alan Greenspan and his army of modellers ought to have caught on to asset bubbles.

    But then I never did understand Greenspan’s cavalier approach to asset bubbles. And he wasn’t alone. I remember at a dinner party in the late 80s asking an eminent international economist about the causes of and prospects for what I casually termed “asset inflation” in Japan, especially the totally insane prices of Japanese real estate. And he was amazingly dismissive, critiqued my terminology, and wandered off into some sort of discourse about random walks. Needless to say, whatever term one wanted to call it or whether it fit into his economic models, Japan had a crazy, crazy asset bubble that eventually burst, with the worst fallout eventually in the corporate sector, even though that wasn’t where the most visible problems were located.

    Anyhow, I start from the old-fashioned assumption, based on the last several centuries of history, that bankers (investors) are lemmings and that financial markets always overshoot (both up and down). The only questions are when and by how much. I think history also teaches us that the most extreme overshooting often involves real estate bubbles because of the time lags involved in getting projects going in response to perceived market opportunities — the latecomers to the party start their projects well after market absoptive capacity will be exceeded by onstream projects but before that exceess capacity has become apparent — and because once the market starts to collapse it then takes a long time to clear excess inventory. Because of this problem of timing and overshooting, there’s an old saw (I haven’t heard in years) that no construction firm or builder worth its salt hasn’t gone bankrupt sometime or another — or to put it another way, the four keys to real estate are location, location, location and TIMING. Maybe that old saw will come back into fashion — ah, the problem of short-term memories and the Great Moderation.

    Given my starting point, to mix metaphors, I look to regulators and central bankers to take away the punch bowl before too many of the lemmings have pranced arm-in-arm drunkenly over the cliff. And if they’re too late to stop the party from getting out of hand, the best the government can do is to use fiscal policy and financial system intervention to help put the brakes on the overshooting on the downside. Which is where we find ourselves today, trying to find a new equilibrium set of asset prices that aren’t the result of a self-feeding deflationary spiral. Or to put it another way, there are many potential new equilibriums, and we want to find one somewhere higher than one that depends on five years of inventory absorption or asset values so decimated that only the vultures remain to clear the market.

    In addition to the costs of the Lehman failure to the real economy which you highlight, Dave, I expect the Lehman shock has made the government’s job of finding a new equliibrium much more challenging, both in that the Lehman shock accelerated the global collapse of asset prices, thereby intensifying the risk of dramatic overshooting in not only the real estate but all other asset markets, and that post-Lehman, Treasury and the Fed find a restricted universe of politicially feasible policy options (and AIG has further radically limited the potential policy universe).

    Now as to where the bond market is going. Whether and to what extent the spread in your chart between AAA and Baa narrows will be a function of what the market interprets AAA and Baa to represent — both what the rating agencies intend to signal via their ratings in the future and how reliable the market believes those signals are. I don’t think the behavior of spreads in your chart reflects a “cooking the books” by the rating agencies — if anything, the current lack of credibility especially of AAA ratings — and the reluctance to downgrade AAA described in the Fons/Partnoy op-ed you linked to — ought to narrow, not widen, the spreads between AAA and Baa. Are you suggesting that Baa has become even more suspect in rating terms (not just less desirable in terms of flight from risk) than AAA since Lehman?

    I do agree with the general thrust of the Fons/Partnoy op-ed. As long as ratings have regulatory consequences for portfolio managers — portfolio limits either in terms of percentage or amount of regulatory capital required — the market assessment of the risk represented by a AAA or a Baa will be distorted. Right now, given the regulatory importance of ratings, the costs of the agencies’ mistakes or biases are enormously magnified, making them as much a source of systemic instability as investor protection. (We should have already learned that lesson in the Asian crisis, but, sigh…) So we’ve got some work to do to figure out how to make information suppliers, like rating agencies, meaningful again without increasing the risk to the system when some information is of poor quality. And I’ve reluctantly come to the same conclusions as Fons/Partnoy — get ratings out of the regulatory system, whether for banks, investment funds, pension funds, insurance, or what have you.

    And while we’re at it, as Fons/Partnoy suggest, let’s also focus on how contractual safeguards, like rating downgrade default triggers, are also contributing to systemic risk. I’d also add to the list of perverse “safeguards” the sorts of legal rules that seem to make sense in the context of a specific transaction or institution, like the bankruptcy preferences for CDS, that turn the weakness of one institution into a system-wide house of cards.

    Insuring against losses produced by bad ratings would just add another regulatory distortion to market information signals, so even if it were feasible, it would be extremely counterproductive. And as your mention of AIG suggests, it’s impossible to imagine we’d find someone with the financial capacity and willingness to put up large enough amounts of “performance bonds”. AIG only got into the crazy business of writing insurance on systemic risk because it didn’t have to put up capital against the insurance it worte — it was monetizing its AAA rating. If it had been required to post collateral and maintain increased reserves, it never would have, nor could have, exposed itself (and the system) to that amount of risk.

    As our financial system has increasingly disintermediated out of banks and into capital markets (where the money center banks have become virtually indistinguishable from the capital markets players), we have been trying to use “safety measures”, like ratings, in lieu of forcing buyers to adequately price the risk they buy or imposing adequate costs on risk-originators. The government is now having to bail out the capital markets (CDS, bondholders of financial institutions) as if they were bank depositors to stop another systemic blowup like Lehman produced.

    If instead of “private insurance” we had been more serious about the government insurance we already provided — that is, if over the past decade the banks had had to pay fees into the FDIC fund — there would have been howls and screams of an unlevel playing field between smaller banks (those that rely heavily on their deposit base) and money center banks (which rely principally on capital markets for funding — direct and off-balance-sheet or contingent liabilities). Instead, we went the politically convenient route, and the banks were allowed to forego FDIC contributions. It’s not that I’m suggesting that payments to the FDIC would have created a big enough fund to have covered the various incarnations of TARP and the Fed’s interventions. There’s no way the financial services industry could pay sufficient premiums for self-funding of insurance against systemic meltdowns. But it would have put a price on deposit funding and forced us to acknowledge that we weren’t pricing capital markets funding and contingent liabilities, which in turn might have forced the issue to be dealt with by regulators and politicians. We might have see some constraints on the obscene amount of leverage being taken on by financial institutions which didn’t have to pay a fully-loaded price for their funding.

    So maybe, since it now appears that the government has to be able to stop “bank runs” in the capital markets as well as in deposits, we need to impose an FDIC-type charge for all sources of financial institution funding, regardless of the type of institution or the funding source. If we really imposed a funding charge on financial institutions — that is, used a pricing mechanism to regulate their liabilities (direct and contingent), not just fret about the safety of their assets a la VaR — we might see a market-driven shift away from mega-bank-monsters and a modest reintermediation of some of the shadow banking system, especially that portion that lives off highly-leveraged, high-volume, high-velocity arbitrage (financial and regulatory).

    I hope the upcoming proposals for regulatory reform reflect these sorts of considerations — removing market distortions that privilege capital markets over intermediation and improving market information and pricing disciplines in the capital markets — rather than focus on old “form” debates like Glass-Steagal or national banking. Even if we imposed extreme “form solutions” such as “narrow banking”, financial crises, like the poor, will always be with us. But we can mitigate the severity of crises and the social problems they produce by putting a bit of sand in the wheels and improving the incentives in the system, not just writing and applying a bunch of thou-shalt-not rules which would be a lawyers full-employment act (she says as a former corporate and securities lawyer) and would stimulate a wave of “financial innovation” in the invention socially-valueless rule-avoidance financial products.

    By the late ’90s (hell, by the S&L crisis), IMHO, the genie of disintermediation and an unsegmented financial system was long out of its bottle, and trying today to stuff it back into a Glass-Steagal form isn’t feasible. But when a decade ago we acknowledged the inevitable by tossing out the old forms, we didn’t take the steps we should have to at least put a handcuff or an anklechain on the genie. Beware of what you wish for…

    By the way, not that you don’t already have too much stuff to read, but you ought to follow the banking industry number crunchers at Institutional Risk Analysis. I don’t always buy their policy recommendations, but their analyses provide lots of food for thought. On the topic of the systemic perversity of “safeguards”, they have today a paper written by a Fed staff member two years ago (and not accepted for publication by the Fed) on pro-cyclical systemic difficulties that might be introduced by a transition to Fair Value Accounting. Now my sympathies for bankers who gamed the system on the upside and don’t want to recognize losses on the downside is pretty thin by now, so I’ve been disgusted by a lot of the anti-Mark-to-Market jeremiads. But the paper makes a number of helpful non-ideological points about the tangle of issues.

    And on the differential treatment of deposits versus capital markets funding — and its differential impact on small and money center banks and the implications for systemic risk — see this fascinating discussion/interview from last week.

Leave a Comment