Paulson’s proposed bailout plan has come under increasing fire, both from the left and the right. While the critics generally are on target, they’re fighting the wrong battle. Paulson’s plan isn’t misguided. It’s FUBAR and it’s time to change the terms of the debate.
What follows is very long and my apologies in advance for that.
I have a dairy yesterday that presents a bit of background on the problem and misconceptions about how we got to our current situation.
http://www.dailykos.com/...
It takes just a bit different spin than Devilstower’s perspective
http://www.dailykos.com/...
Many top analysts have unloaded on Paulson’s proposal, for example, Robert Reich,
http://tpmcafe.talkingpointsmemo.com...
Sebastian Mallaby,
http://www.brookings.edu/...
Willem Buiter,
http://blogs.ft.com/...
and Brad deLong
http://econ161.berkeley.edu/...
While each of these are good extensions or modifications of Paulson’s proposal, they are fundamentally flawed in one dimension. They use Paulson’s approach as the starting point. From the perspective of what comes out of Congress, that’s reasonable and understandable. However, that also puts you on the most tilted playing field imaginable.
So let’s take this in two steps. First, how did we get into this mess? And second, what do we need to do to get out? Taking Paulson’s plan
http://www.nytimes.com/...
as the starting point is worse than doing nothing. It doesn’t address the fundamental causes of the problem, and by doing something stupid we paper over the problem and make a true solution less likely.
Genesis - How we got here.
(1) Lack of Understanding. Drexel Burnham started collateralized debt obligations (CDOs) back in 1987. There is nothing, per se, wrong with CDOs. To the extent that they allow the seller to repackage debt and sell it to buyers may make everyone better off. The seller gets the debt off its books while the buyer gets an obligation with an income stream and risk characteristics that fits its needs. An insurance company that has a premium stream coming in and future expected payout requirements can match the expected payout requirements with the expected income from the security. That’s assuming that buyer and seller understand the behavior and characteristics of the security.
Mortgage backed securities (MBSs) date back even further and have had a transformational effect on the housing market. Banks can issue mortgages and not worry about having loans concentrated in a particular locale, thus reducing bank risk, and mortgages aren’t limited to the amount of funds that can be raised locally.
The bottom line is that CDOs and MBSs and other acronyms and wonders of financial innovation can dramatically improved our economic well being, both by channeling funds where they do the most good and by reducing risk or allowing institutions to match inflow and outflow streams. However, the positive contribution of CDOs and MBSs are predicated on both parties to the transaction understanding exactly the nature - risks and returns - of the derivative security and its underlying assets.
Lately, the lack of knowledge has been a critical part of the problem. Once you have your garden variety MBSs, if you’re at Drexel or Lehman or Goldman and you want to increase profits, then you need to be more creative. That’s not to knock creativity but there are limits. Once you have three blades on a razor, how much additional benefit will you get adding a fourth or fifth or ...? But if I’m at Drexel and you’re at Goldman and I want a competitive advantage, I’m going to see if a CDO of MBSs might work a little better or entice a bit more business than just an MBS - so I’m going to add a ninth blade to the razor.
Very quickly you can get to a world where only the financial engineer of a particular instrument understands what’s going on. It’s a "new and improved" version of what the sales guys have been pushing so out it goes - and out goes the understanding of the characteristics of the instrument - just at the time when those characteristics are becoming increasingly hard to estimate.
If everyone could value the myriad securities on offer, we simply wouldn’t have a liquidity crisis! The prices of many CDOs might be going through the floor and some firms might be going bankrupt, but there would be no liquidity crisis.
Now for the technical types out there, consider two additional issues. (a) What is the probability distribution of returns of a particular security? In particular, can you define precisely what is in the tails of the probability distribution, e.g. what’s the probability of a complete meltdown in any number of dimensions? Our financial models do a poor job of estimating these probabilities because we simply don’t have enough observations to quantify the risks. The failure of LTCM is a classic example. The economics Nobel winners running LTCM mis-estimated the risks in the tail of the distribution and their quant-based firm was toast. (b) What is the correlation between financial instruments? If you’re going to create a financial instrument that will reduce one type of risk, you want to pool assets with different characteristics, e.g. SUV auto loan values may fall when oil futures prices rise. When you combine those in a single security, you get less risk - under your assumed scenario. However, a different scenario may lead to a very different outcome. For example, if we move into a severe recession, the demand for SUVs and oil both decline and both prices drop. In other words, the correlation between the assets has gone from negative - which is what you were looking for to reduce risk - to positive with the latter potentially increasing risk.
Part one of the process: lack of clarity and lack of understanding of financial instruments. Now you might argue that there are armies of Ph.D.s working for the Drexels and Goldmans. Lack of understanding shouldn’t be an issue. And you would be correct. It shouldn’t be. However, the lack of transparency in the valuation is a large component of what has shut down trading in financial markets. Drexel and LTCM in some sense shouldn’t have failed given their abundance of very smart people. But fail they did.
(2) Lack of Appropriate Incentives. "Greed is good." That view, with a small twist, underlies our economic system. Virtually any economic model starts with the assumption that firms try to maximize profits and then makes decisions consistent with that objective. So pricing, employment, product mix decisions are all driven by efforts to generate the maximum profits. Profit maximization is a polite way of stating greed.
Now few firms actually have the opportunity to maximize profits without constraints, e.g. legal constraints, environmental constraints, capacity constraints, contract constraints, regulatory constraints, ... You can think of your typical profit-driven firm as pushing up against myriad constraints. And your typical firm may also be doing what it can do relax those constraints, e.g. regulatory constraints.
Well, if greed is good - or if constrained profit maximization is good - then what’s the problem? The simple answer is in how it’s rewarded. Economists would like to think that CEO performance was valued based success in increasing long-term profits. However, far too often this perspective has changed to "What have you done for me lately?" Did you beat last quarter? Did you beat last year? Did you meet expectations? How much growth did you achieve this month or this quarter or this year? The CEOs - and many others - in Lehman, Goldman, Merrill, ... have pulled down astronomical pay packages because their firms’ performance last year was good. If I can boost profits for the next couple of years and make $25M or $50M or $100M or more, do I really care that I’ve incurred tremendous additional risks that might bankrupt the firm in the future? Some CEOs will say "my reputation matters" and will add that as a constraint when maximizing; others will simply take the money and let someone else worry about issues arising later. Part one of the incentive problem is a system-wide focus on the short run.
Part two of the incentive problem is a desire to avoid admitting mistakes. By summer 2007, the housing market was clearly heading south. Anyone holding MBSs or CDOs based on MBSs should have recognized that there were major problems and should have increased their loan loss provisions dramatically. A few did; many did not. Doing so early in the game would have required admitting mistakes and would have lowered short-term profits. However, it also would have allowed earlier recapitalization if necessary.
Part three of the incentive problem is the issue of moral hazard. What has happened with financial market failures? The most recent examples: LTCM was bailed out with a shotgun marriage as was Bear Stearns. A.I.G. received similar treatment although stockholders may yet raise sufficient capital to buy the company back. While Lehman was allowed to fail, many high-profile employees are being aggressively pursued by headhunters and have moved or are moving into very well compensated positions. My point: those who have made serious mistakes often have been largely shielded from the effects of their mistakes. The principals at Bear, A.I.G., and Lehman will live extremely well for the rest of their lives even if they never work another day. Best case scenario for these decision-makers: they’re modern-day Midas’s. Worst case scenario: they’re only extremely wealthy, albeit perhaps with a tarnished reputation, and they get early retirement.
(3) Lack of Regulation. Deregulation fundamentally started under President Carter with the airlines. (The Ford administration initially pushed it ineffectively, and the Reagan administration turned deregulation into a mantra.) And deregulation can be a key feature to stimulate growth, innovation and development, as it has been in the airline industry. Nevertheless, regardless of your political views on regulation, you should go back and read Adam Smith. He recognized the necessity of regulation to ensure the integrity of the marketplace. The wholesale push to deregulate financial services has threatened that integrity.
Make no mistake. I am not supporting wholesale government regulation or re-regulation. What I am supporting is government setting out the ground rules for the game. Devilstower’s piece
http://www.dailykos.com/...
spells out the details of deregulation and how that has impacted financial markets.
Markets have imperfections - limited information for example - and those imperfections are a reason why some regulation is necessary, even for the very integrity of the marketplace. Stock markets, for example, have regulations about who can sell (e.g. no trading on inside information), when you can trade (e.g. the uptick rule through July 2007), and what you must disclose (e.g. if you amass or attempt to amass more than a certain percentage of a firm’s stock). Those regulations are designed to make stock trading a fair game for all participants and by so doing encourage greater trading. Finance professors might want to tweak the regulations but they generally don’t argue against eliminating all of them.
I spent some time in Russia in the early 1990s. As an economist it was a bit puzzling and frustrating that they were much less interested in economic issues than in legal issues. Property rights were their key interest and I didn’t find that interesting at all. In retrospect, their interests were on the mark. Whether you have the Russian system of oligarchs and a quasi-governmental system of markets or whether you truly have a competitive system depends on how you structure your legal rules to the game. To the extent that you adopt a mantra of "deregulate, deregulate, deregulate" then you get to a point where you have no rules of the game - or you’re playing Calvinball. A competitive marketplace is not consistent with Calvinball and that certainly appears the direction in which we have moved.
Some of my colleagues contend that the Community Reinvestment Act’s (CRA) 1995 modifications prompted the current liquidity crisis. That is, loans made to low-income borrowers under CRA generated higher defaults that sent the housing market and the MBS market into decline. No evidence supports that contention. These loans were initially made starting in 1995. If greater problems originated there, we should have seen them developing long before now. The recent upturn in defaults is clearly higher among CRA loans but is seen in all classifications including conventional and jumbo.
In contrast to bank lending, the market for CDOs is basically the Wild West. The creative minds at Goldman or Morgan or Merrill cook up the CDOs. Their marketing people sell them with virtually no oversight or regulation. If everyone in the market is fully informed, there’s no problem. When some are not fully informed, clearly they shouldn’t be in the market. But it’s these rubes that Paulson’s bailout will most directly help.
Bottom line: regulation may help or hurt markets depending on how it’s structured, and the financial market most directly impacted is one where regulation is sorely lacking.
How we avoid getting here again - or don’t forget the S&L debacle again
Paulson’s bailout is not the solution. It doesn’t address the proximate problem, a drop in housing values. It addresses only a symptom, a drop in the value of securities based on housing values. And it doesn’t address the fundamental underlying problems, lack of understanding of how some of the securities behave, lack of proper incentives for financial CEOs in particular, and lack of regulation to protect consumers and taxpayers from excesses due to the first two problems. Paulson’s bailout potentially makes matters worse rather than better by totally ignoring the first two problems and by pushing the third into the future when many will be tempted to say the problem has been solved and shift back to the "regulation is unnecessary" argument. Mallaby and deLong and others do an admirable job suggesting clean-ups to Paulson’s proposal, but they’re really trying to slow a runaway train when they don’t have access to the brakes or the throttle. (I don’t either but my goal is to break down the door to those controls!)
The problem as outlined above has three parts and thus the solution has three parts. Let me stress at the outset that these are long-term issues with long-term solutions.
(1) Knowledge. There’s been much work, in particular by the SEC, to require plain language rather than boilerplate in financial filings like 10-Ks. Those efforts need to extend to all financial instruments. Personally, I would adopt what some in the industry might consider a draconian standard: if you can’t clearly explain your derivative security - both its composition and why it represents value added - in a page or less then you don’t have a worthwhile product and you will not be allowed to sell it. In addition, there should be a clear statement of the assumed probability distributions, payouts, and correlations underlying the security. MBSs generally would pass this test easily; most of the toxic CDOs would not. It will take additional regulation to enact this provision but it should be a no-brainer. Financial innovation wouldn’t be eliminated but it would no longer be an end in itself. If Wall Street can structure new products to help Main Street, then the product should be able to pass this test. Otherwise, no new derivatives.
(2) Incentives. There’s an extensive economics literature on executive compensation. How should compensation be paid to align management interests with shareholder interests? Efforts to compensate management with more stock options and less salary largely stem from that literature. I’m not going to try and summarize that literature here. I will only suggest a change that would partially alleviate the current problem. Financial firms can set any payoffs they want just as they currently do. However, payoffs in excess of some amount ($0.5M/year?) are sequestered for a minimum of five years. (I would prefer ten.) That is, if you make $5M this year, you immediately receive $0.5M and the remainder only later and only if the firm doesn’t go bankrupt. Executives could borrow against the funds and could earn interest on the funds. However, in the event of default any prior executive pay gets placed in bankruptcy proceedings and the executives get treated like any other ordinary claim holder. This change would force executives to focus more on the long run and would reduce some of the speculative excess that we’ve seen in the past few years. Lehman’s top five earned $81M in 2007 and they may well be sued by stockholders under bankruptcy law. This proposal dramatically increases the costs to executives of a bankruptcy since five years of virtually all pay would be at risk rather than one year. Additional provisions should be added to address pay issues with "shotgun weddings" such as Bear Stearns. Logically, CEOs should take the same haircut as shareholders in these circumstances. The goal with this change is to reduce moral hazard. If you’re an executive of a firm that "requires" a bailout or simply goes bankrupt and destroys value, you bear much of the blame and deserve a severe financial penalty
(3) Regulation. I don’t want to try and write specific regulations at this point. I’m not sure that I or just about anyone else is in a position to do so. However, if you’re going to claim that your firm is integral to the economic health of the country or your firm going out of business is going to precipitate a credit crisis, then you’ve forfeited all ability to say "I shouldn’t be regulated."
Let me state this point as absolutely explicitly as possible. Paulson’s proposal is for $700B; adding funds already committed to AIG and Fannie and Freddie and Bear yields a sum of roughly $1 Trillion. With a population of about 300 million, that means we will on average pay $3,333 for this overall bailout. Wall Street lobbyists have been pushing Republican lawmakers to pass a clean bill with no strings attached meaning no regulations. Excuse me but if Paulson or Bush or Bernanke or Goldman or Merrill or Citi or whoever ask you for $3,000 and change, do they seriously expect that you’re going to simply say "yes?" Regulations are a sine qua non even for consideration of a bailout. And the regulatory hurdles added should not stop with MBSs and CDOs but need to add Credit Default Swaps (CDSs) as well, a market that dwarfs even the market for mortgages and which I consider largely speculative (as does Warren Buffet who called them "financial weapons of mass destruction"). (For completeness, with a CDS I would pay a seller a fixed payment basically for insurance against a third party defaulting. Thus, if I held a Bear Stearns bond that defaulted and had a CDS with A.I.G. on that bond, then A.I.G. would make up the shortfall on the defaulting bond.)
The marketplace for MBSs, CDOs and CDSs is basically the Wild West and nearly anything goes. There is little regulation, especially for CDSs. In many respects it’s similar to eBay except less organized. If I put something up for sale and if you are the high bidder, it’s yours. There is little standardization. For MBSs, there has been some liquidity for quite some time with Fannie Mae MBSs serving as the benchmark. You need some standardization in order to be able to have some liquidity and to facilitate trading. For the CDOs and CDSs, there is so much variability in terms of the assets underlying the contracts and in terms of the credit worthiness of the sellers that the market generally has been "thin." That is, if you purchased a CDO from Lehman, for example, your options other than selling it back to Lehman are very limited.
For years Republican congressmen railed against Fannie Mae and Freddie Mac for being undercapitalized. For the last year, it would be hard to argue with them. However, requiring greater capitalization is one component of regulation. Fannie and Freddie are likely not alone with capitalization issues and I suspect that the lack of sufficient capitalization of firms issuing CDOs and CDSs may have spooked Paulson into his bailout proposal. However, we don’t even know how many CDOs and CDSs are outstanding. The best we have are estimates of the CDS market size of over $45 trillion for example. (In contrast, the Treasury market value is about $5 trillion and the total U.S. stock market value is about $20 trillion.) Thus the first step in the regulatory process is for all issuers of CDOs and CDSs to report their issuance or offers to an agency like the Federal Reserve. The second step is to require adequate capitalization for all issuers. It makes little sense to regulate A.I.G.’s insurance arm when their CDS unit is operating with no regulation and bringing the entire firm to the brink of bankruptcy requiring government assistance. A third regulatory component would be to attempt to provide some standardization of CDOs and CDSs. That’s a small step, I believe, compared to the first two. However, if the goal is to provide liquidity, lack of standardization is a major impediment and this step may be welcomed by the market.
What REALLY is the problem?
This discussion at some points may appear completely divorced from what Paulson proposed. That should come as no surprise. What he proposed was far removed from solving the problems that the U.S. faces. His proposal appears structured to deal with a short run liquidity issue. Dealing with that short run issue has crushed Fannie and Freddie and Bear and Lehman and Merrill. Each tried to restructure their assets to provide liquidity to meet what they thought were temporary shortfalls. None addressed the underlying issue. None had the ability to address the underlying issue.
That underlying issue isn’t liquidity. From some financial institutions’ perspectives, it’s solvency. Property values have decreased. Thus, any financial instrument based directly or indirectly on property values will also have decreased. Buying bonds from financial institutions at "fair market values" only makes it clearer that their problem isn’t liquidity. Selling at fair market values means those financial institutions can and must mark their assets to market and must recognize capital losses that have been hidden thus far. Only if the Treasury buys vast amounts of illiquid assets at below market prices will their problems be solved. Well, actually their problems then become our problems!
Even the compromise with additional oversight does little to address the underlying issue. Property values have declined and thus the value of financial instruments based on them have declined. If you want to increase the value of those financial instruments, fundamentally you need housing values to rebound. Steps like limiting foreclosures or restructuring mortgages will help.