Reforming the Financial System
May 17, 2010

The financial crisis and its lingering aftermath have revealed serious vulnerabilities in our financial system. The challenges are comparable to those posed by the great depression of the 1930s. Yet the approaches taken to date to fix the problems have been fitful, fragmented and frenzied, rather than deliberate and thoughtful. That’s what makes me think we need a financial commission to sort things out.  

The Case For a Financial Commission

Of course, the political process is most receptive to reform while memories of the crisis remain strong and its after-effects are still being felt. But there is a serious downside to legislating reform while we are so close to what happened. Reform proposals tend to focus on preventing the same sequence of events that we very recently lived through. This is reminiscent of the old adage about generals preparing to fight the last war. The next potential financial crisis is going to be different from the last one, just as that one was very different from the savings and loan crisis of the early 80s.

There is another very serious drawback to legislating reforms while memories of what happened are still fresh. Many of the proposals are suffused with hostility toward groups that are viewed as malefactors – especially “bankers” and “Wall Street”, which have almost become curse words. It’s easy to be mad, especially if you have been burned by the crisis in some way, but it is not a good backdrop for the thoughtful legislative reforms that we need.

Nothing disrupts the process of rational thought quite so much as the prior conviction that the problems were caused by the greed of wicked people. But the fact is that commercial bankers, investment bankers and mortgage bankers were not suddenly caught up in an epidemic of greed, they have always been greedy. Usually, their greed has acceptable, even good results for society but occasionally a business and regulatory environment arises in which it leads to disaster. If we intend to prevent another crisis, that environment should be the focus.  

 To examine the issue properly requires a coherent approach that is most appropriate to a commission with a broad mandate. In the best of all worlds, the findings of such a commission would precede legislative proposals. The remainder of this article consists of agenda suggestions for such a commission. They are organized in terms of the following six major issues:

     *Is There a Way to Allow the Failure of Firms That Are Now Viewed as “Too Big to Fail” (TBTF)?


     *Why Have Regulators Failed to Prevent Excessive Risk-Taking by TBTF Firms?

     *If Capital Requirements Can’t Do the Job, What Can?

     *How Should We Redesign the Home Mortgage System?

     *What Should Be Done With Fannie Mae and Freddie Mac?


     Is There a Way to Allow the Failure of Firms That Are Now Viewed as “Too Big to Fail” (TBTF)?

The most traumatic and disruptive feature of the recent financial crisis was that the Government was forced to rescue firms that had behaved recklessly. These firms were “too big to fail”, meaning that the repercussions of their failure would have destabilized the entire system. The best analysis of this problem that I have seen is in Wind-down Plans as an Alternative to Bailouts: The Cross-Border Challenges, by Richard Herring, a Wharton colleague. I have drawn heavily from his paper.

TBTF firms in some cases provide services that are critical to the functioning of markets, such as acting as a dealer or providing settlement or escrow services. In some cases, they own many hundreds of affiliates in foreign countries, failure of which become problems for each country’s regulators, who may or may not talk to each other. TBTF firms also are likely to have credit and other obligations in such large volume and to so many other firms that a failure to pay could panic investors and cause markets to freeze.

The need to rescue TBTF firms was underscored by the Lehman Brothers case, where officials could not find a legal way to effect a rescue in time; the ensuing disruption was devastating. Herring notes that “Lehman’s bankruptcy has led to civil proceedings on three continents where transactions were aborted in the middle of the clearing and settlement process.” He notes that about forty-three thousand transactions are still open and have yet to be negotiated or litigated.

The unstated policy adopted after the Lehman failure was that no more Lehmans would be permitted because the costs were excessive. But the cost of rescues is also staggeringly high.

The immediate cost of rescue operations is the expenditure of public funds to protect the creditors of TBTF firms. The long-run cost is that the rescues of TBTF firms that have already occurred increase the likelihood that even more costly rescues will be needed in the future.

The high probability of rescue gives firms an incentive to become TBTF if they are not one already. TBTF firms enjoy a competitive advantage based not on their efficiency or customer service, but simply on their being TBTF. Since their creditors are confident that they will be protected, there is no need for creditors to monitor the soundness of the TBTF firms to whom they entrust their money, which encourages TBTF firms to earn more by taking more risks. In short, if the problem is not fixed, it will get worse.

Nothing that has been proposed by the Administration or the Congress to date will eliminate TBTF firms. Size restrictions will probably be adopted, but will have little impact unless they require widespread downsizing, which nobody is proposing. The proposal for a Governmental resolution authority that, in Paul Volcker’s words “should be authorized to intervene in the event that a systemically critical capital market institution is on the brink of failure,” might make the rescue process more orderly and less frenzied, but it won’t stop rescues. On the contrary, it will probably make rescue more certain by eliminating the possibility that Government won’t be prepared to take action. Had such an authority been involved in the Lehman case, it is almost certain that Lehman would have been rescued.

A Government resolution authority would have more muscle if it could require that all systemically important firms develop an approved “wind-down plan”, which would specify exactly what the firm would do in the event that it became insolvent. Herring has a detailed description of exactly what a wind-down plan should include. To be approved by regulators, the plan would have to show how the firm would wind up its affairs, and those of all of its affiliates, in a reasonable time frame and without any adverse spillovers to other firms.

This approach would be fiercely resisted by TBTF firms and would be extremely challenging for regulators to implement effectively. If successful, wind-down plans would eliminate or reduce disruption costs, such as those that result from the aborting of transactions in process, but it is not at all clear that it would eliminate the perceived need to protect creditors. A wind-down plan won’t necessarily prevent a contagious loss of confidence if the firm doesn’t pay its debts,

It follows that equal if not greater emphasis ought to be given to preventing TBTF firms from getting into trouble in the first place.  

Why Have Regulators Not Prevented Excessive Risk-Taking by TBTF Firms?


For reasons discussed above, policies designed to convert TBTF firms into firms that are not TBTF are unlikely to be enacted. The alternative to rescuing such firms is to prevent them from getting into trouble in the first place by effectively regulating their risk exposures.

But our regulatory history is not encouraging. The savings and loan industry, rocked by a different kind of crisis in the early 80s, was regulated by the Federal Home Loan Bank System (FHLBS), which Congress subsequently terminated because of its regulatory failures. Some of the major banks and thrifts caught up in the recent crisis were also subject to extensive regulation. The Citibank and Bank of America holding companies were regulated by the Federal Reserve while the banks themselves were regulated by the Comptroller of the Currency. WAMU, Indy Mac and Countrywide were regulated by the Office of Thrift Supervision, successor to the FHLBS, and it will probably be terminated in turn for its regulatory failures.

If we expect regulators to do better next time, we better understand why they struck out this time. Part of the reason is that major players contributing importantly to the crisis were, for all practical purposes, unregulated. These include investment banks, mortgage companies, and AGI, an insurance holding company.

It is very difficult to rein in regulated institutions that must compete with unregulated firms, or are seduced into making attractive deals with unregulated firms. Furthermore, regulators tend to be jurisdictionally myopic; they ignore threats to financial stability that arise outside of the industry of firms over which they have legal jurisdiction.

Conclusion: somehow the regulatory net must be made wide enough to cover all the firms that become important players in the next emerging bubble that can lead to a crisis. This is a major challenge, since the important players next time are unlikely to be the same as the important players in the most recent crisis.  

The second cause of regulatory failure is the lack of adequate regulatory tools. The principle tool on which regulators rely to prevent excessive risk-taking is capital requirements. This is plausible because capital requirements focus directly on a firm’s capacity to bear loss. If a firm has a capital ratio of 4%, for example, it means that the value of its assets could fall by no more than 4% before it is insolvent. If an 8% ratio is required, its assets must fall by 8% before it is insolvent, which means it is stronger.

But capital requirements haven’t worked. For one thing, the amount of capital a firm has to cover losses can change sharply within a very short period as a result of asset value write-downs. Richard Herring notes that ”The five largest US financial institutions that either failed or were forced into government-assisted mergers in 2008 – Bear Stearns, Washington Mutual, Lehman Brothers, Wachovia, and Merrill Lynch – were each subject to Basel capital standards, and each disclosed Tier 1 capital ratios ranging from 8.3% to 11% in the last quarterly report before they were effectively shut down.” (Wind-down Plans As An Alternative to Bailouts, 2010).  

Furthermore, capital requirements don’t discourage risk-taking during a bubble because firms can shift the composition of their assets toward greater risk without increasing their required capital. For example, the shift from prime into sub-prime mortgages did not increase capital requirements because both belonged to the same broad asset class.

Capital requirements do not increase the capacity of firms to absorb loss during a bubble unless the regulators increase the requirements. But this requires special insight into the economy by regulators, not to mention the political courage needed to remove the punch bowl from the party. This is too much to expect of regulators.

But perhaps the most serious limitation of capital requirements is that they only apply to firms in the industries to which the requirements have been applied. These are the industries that have been implicated in disturbances in the past, whereas the disturbances of the future may come from different sources.

The challenge is to find a regulatory tool that will a) automatically discourage risk-taking during a bubble, b) increase the capacity to bear loss of firms participating in the bubble, and c) apply to any systemically-important firm, regardless of what industry they happen to be in.  

Preventing Excessive Risk-Taking by TBTF Firms

One of the few useful proposals that has emerged from the post-crisis post mortems has been to create a systemic risk regulator whose jurisdiction is not limited to conventional industry boundaries. It is essential that the regulatory net be made wide enough to cover all the firms that might become important players in the next emerging bubble leading to a crisis.

It is also essential that the systemic risk regulator have the proper tools. Capital requirements, which regulators now depend on to ensure safety and soundness, don’t do the job, for reasons noted above. The regulator must be able to remove some of the profit from taking excessive risk during a bubble period, while requiring that firms taking on these risks increase their capacity to bear loss.

Transaction-Based Reserving

The needed tool is transaction-based reserving, or TBR. Under TBR, financial firms are obliged to contribute a part of risk-based income to a contingency reserve account that is not accessible for 10 years except in an emergency.  Income allocated to reserves would not be taxable until it was withdrawn 10 years later. 

Here is an over-simplified example. The lender makes a prime home mortgage loan at 5%, the risk component of the rate is 1%, and the TBR is half of that or 0.5%. The lender shifts to a sub-prime loan at 7%, the risk component is now 3% and the TBR is 1.5%.  The capital requirement doesn’t change, but the TBR reduces the profitability of the shift, and if the lender does it anyway, the required allocation to the contingency reserve becomes three times as large.

We have extensive experience with TBR in connection with private mortgage insurance companies (PMIs), which have been subjected to it since the industry began in 1956.  PMIs have allocated 50% of their premium income to a contingency reserve for 10 years. These reserves have allowed the PMIs to meet all their obligations in connection with the extraordinary losses suffered by lenders during the current crisis. They may or may not make it, depending on how long the crush of foreclosures lasts, but if not for TBR, they would have received Government support or disappeared a long time ago.

A systemic risk regulator armed with TBR could cope with the next threat to financial stability, whatever that turned out to be.  Because TBR is transaction-based rather than institution-based, the regulator’s jurisdiction would extend to any institutions involved in transactions that it viewed as a threat to systemic stability.

Of course, the rules could vary by type of institution, and lots of details need to be worked out, such as the rules for determining the risk component of various kinds of transactions, and for investing and safeguarding reserves. Rules must also be developed for firms that originate risky transactions and then sell all or part of their interest. In such cases, the required reserve allocation might be divided among all those in the chain of ownership, which would be a way to assure that they all had some “skin in the game.” Alternatively, the entire allocation might be imposed on the last actor in the chain, who would transmit the burden back through the chain in the prices paid.

A great advantage of TBR, relative to capital requirements is that TBR does not depend on discretionary actions by the regulator to offset the excessive optimism that feeds bubbles. Once the TBR rules are established, a shift to riskier loans during periods of euphoria automatically generates larger reserve allocations because riskier loans carry higher risk premiums.

Redesigning the Home Mortgage System


The housing finance system of the US, once viewed by many as a model, is now a shambles. The underwriting rules applicable to 9 of every 10 mortgage loans, stipulating who is and who is not eligible for the loan, are dictated by an arm of the Federal Government: Fannie Mae, Freddie Mac, FHA, VA and USDA. The sliver of the market not touched by those agencies is dominated by a small group of too-large-to-fail bank holding companies.   

Before the crisis, the non-Federalized part of the market was much larger because it was supported by a private secondary market. Lenders originating loans that did not meet the requirements of any of the Federal agencies could sell them in the private secondary market.  But that market collapsed in 2007 and has yet to reopen.  

One result has been a sharp widening of the yield spread between conventional loans that can be sold to Fannie Mae and Freddie Mac, and those that can’t. Today, the spread between a $417,000 loan purchasable by the agencies and a $418,000 loan that isn’t purchasable by them, but which is otherwise identical, is almost 1%. Before the crisis, it was 1/4-3/8%.

The key to an effective redesign of the housing finance system is the development of an effective private secondary market, but not the kind of market we had before. That market was markedly inferior to the Danish model, which could easily be transplanted here.

Structural Instability: The US market had a structural defect that made it extremely vulnerable to a contagious loss of confidence. Every individual mortgage security was a stand-alone entity secured by whatever reserves or insurance protections were embedded in that security. If these reserves turned out to be superfluous, as they always were before the crisis, they were paid out to investors who owned a residual claim to them. These were often the firms that had issued the security. While these firms had the right to remove unneeded reserves, they were under no obligation to provide additional reserves if this proved necessary.

Since the surpluses on one security were not available to meet deficiencies on others, and since no one was obliged to provide additional reserves if this became necessary, the entire market was a house of cards. When some securities did run into trouble and were downgraded by the credit rating agencies, fears about the status of others ran rampant and the entire house collapsed.

In the Danish model, in contrast, every mortgage security is a bond that is a liability of the firm issuing it. The Danish mortgage banks issue multiple bonds, and if one of them experiences a high loss rate, all the resources of the bank are available to deal with it. There has never been a default on a Danish mortgage bond in over 200 years. During 2008 when the mortgage-backed security market in the US collapsed and the mortgage bond market in continental Europe froze, it was business as usual in the Danish mortgage bond market.

Linkages to Primary Markets: In the US model, the secondary market  and the primary market where loans are made to borrowers, are distinct, connected only through transfers of ownership over a period of time. For example, a loan closed by a small (“correspondent”) lender is sold to a larger wholesale lender who sells it to an investment bank who places it in a new mortgage security. Months may pass between the date when the loan is closed and the date when the loan becomes collateral for a security.

In the Danish model, in contrast, there are no transfers of ownership, because each individual borrower is funded directly by the secondary market. The mortgage bank sells the mortgage to investors simply by adding it to an open bond issue covering the same type of mortgage. If the new loan is a 5% 30-year FRM, for example, it is added to the outstanding bond secured by 5% 30-year FRMs.

Reflecting these differences in the relationship between primary and secondary markets, borrowers in the US face far more challenges in shopping for mortgages than borrowers in Denmark. Borrowers in the US don’t have access to secondary market prices, and if they did, it would do them no good because there would be no way to use it. They are on their own in dealing with loan originators, many of which use a variety  of tricks of the trade to extract as much from them as possible.   

In Denmark, borrowers can price their loan by accessing secondary market prices on-line. They enter the type of mortgage they want and the interest rate, and find the corresponding bond selling for the highest price. The prices of all Danish mortgage bonds are shown on the NASDAQ web site,  (Alternatively, they can go to a broker or loan officer who is paid by the lender selected, who has access to the same bond data with consumer-friendly add-ons.) The borrower pays the bond price plus a .5% rate add-on by the lending bank, plus some out-of-pocket fees that are set competitively. 

Refinancing Options: When market interest rates drop, borrowers in both the US and Denmark, can refinance at par to lower their interest rate. When market interest rates rise, however, only borrowers in Denmark can refinance at the lower market price. Borrowers in the US must pay off their old loan at par.

For example, Doe has a $200,000 balance on his 5% mortgage, and he expects to sell his house for $250,000 in a market in which home buyers pay 5%. But before he can sell, market rates jump from 5% to 7.5% and potential buyers can now only afford to pay $200,000, wiping out Doe’s home equity. However, because of the rate increase, the market price of Doe’s 5% mortgage has dropped from 100 to 85. If Doe is a Dane, he can refinance into a 7.5% loan by paying $170,000 to retire his old loan; by so-doing, he retains 3/5ths of his equity. If Doe is from the US, his entire equity is wiped out.

Given the already substantial depletion of home equity in the US, the need to reduce the further losses that will occur when interest rates begin their inevitable ascent, is compelling.

Note: I am indebted to Alan Boyce, the only person I know who understands the Danish system better than the Danes. 

What Should Be Done With Fannie Mae and Freddie Mac?


The Administration does not know what to do with Fannie Mae and Freddie Mac. While the mixed private/public model under which they had operated has been thoroughly discredited, and the agencies are now in a Government conservatorship, they are critically important in today’s market. Because there is a fear of rocking the boat, the issue of what to do with them going forward has been placed on hold. Such delay is a good thing if the time is used to get it right.

In the long-run, one of the agencies could be entrusted with developing a private secondary market to replace the one that collapsed during the crisis. In the short  run, both agencies need to get on-board the Government’s policy of stimulating economic recovery.

Developing a New Private Secondary Market: As discussed above, the new market should have the following major features:

*Firms issuing mortgage securities retain full liability for every security they issue. This aligns the interests of loan originators and investors, which makes the market safe for investors.

*New loans are financed by selling them directly into open security issues of the same type. This makes the market efficient.

*Borrowers have direct access to the secondary market, borrowing at the price of the security sold to finance their loan, plus a rate markup and fees charged by the lender. This makes the market transparent, protecting borrowers against being over-charged.

*Borrowers have the right to pay off their mortgage by buying back an equivalent amount of bonds, at the lower of par and the market price. This makes the market stable, protecting borrowers against rising market interest rates.

Fannie Mae and Freddie Mac should compete to determine which will be selected to develop this new market. The agency selected would be the one with the best implementation plan as determined by a special inter-agency group appointed for the purpose. The selected agency should recast its own secondary market operations so that the markets work in the same way.

The other agency should be placed in liquidation. If there ever was a need for two agencies in the past, there clearly will be no such need in the future.

Participating in Economic Recovery:  Although they are now part of the Federal Government, the agencies are operating at cross-purposes to the Federal Reserve, as if they are trying to earn their way out of conservatorship. While the Federal Reserve has been purchasing huge amounts of their mortgage backed securities to lower interest rates for home buyers and refinancers, the agencies have been adding a series of price add-ons that have raised financing costs to countless millions of potential borrowers.

The price increments imposed by the agencies are called “Loan Level Pricing Adjustments” (LLPAs) in Fannie jargon, and “Post-Settlement Delivery Fees” (PSDFs) in Freddie jargon. They are upfront payments expressed as a percent of the loan balance and are based on loan-to-value ratio (LTV), FICO credit score, and various loan characteristics related to risk. Before the crisis, they were largely limited to special programs, but in March, 2008, they were applied to standard programs. In March of this year, they were increased by both agencies.

For example, the LLPA today on a loan to a borrower with a credit score below 620 is 1.5% if the LTV is 60.01-70%, and 3% if the LTV is above 70%. If the loan is on a property with 2-4 dwelling units, there is an additional LLPA of 1%. If the property is a rental, another LLPA kicks in for 1.75% if the LTV is 75% or lower, 3% if the LTV is 75.01-80%, and 3.75% on LTVs above 80%.

LLPAs and PSDFs are cumulative. Fannie Mae will charge a borrower with a FICO below 620, an LTV of 80.1-85%, buying or refinancing a 2-unit investment property 3.00% + 3.75% + 1% = 7.75%. More likely, such a borrower would give it up, as millions have.

I am a strong believer in risk-based pricing, but the agencies are over-shooting the mark in order to generate revenue, as if this will offset their excessive liberalizations during the years of boom and euphoria. If a borrower applying to refinance has been current on his loan for at least the last 12 months, there is no good reason to charge any price premium at all, regardless of LTV, credit score, type of property or loan purpose.

Underwriting requirements have also been tightened, with appraisal rules becoming a major bottleneck.  On May 1, 2009, the agencies issued a Home Valuation Code of Conduct (HVCC), which is discussed in The Appraisal Debacle: An Example of How Not to Regulate. The upshot is that HVCC has reinforced the downward bias in appraisals that developed after the crisis, increased the time it takes to get an appraisal, eliminated the ability of a borrower to use the same appraisal with multiple loan providers, and made it impossible to get any appraisal at all when the appraiser can’t find three comparables in the same area. Before HVCC, I never ran into cases where prospective borrowers were unable to get any appraisal, but now I see them often.

The agencies should be required to find ways to undo the damage that HVCC has inflicted on this market.

Sign up to Receive New Articles