In late 2008, our Wall Street Insider, a trader with a decade of experience in the world of structured finance, wrote a trader’s account of the underlying causes of the financial crisis aimed at a general readership. The technical details, though still complex, are boiled down in a way that is understandable, and in a way that points to the broader absurdity of this esoteric world. This version is a bit dated at this point (a more current version can be found here) but contains a bit more detail about how the trading desks for securitized products actually work.

A Brief History of the Financial Crisis

The financial crisis today is like a box of chocolates. Or maybe more like an onion. There are a lot of layers, and many different ways in which to understand it. I guess it makes more sense to focus on the goal, rather than means, of analysis. There are a few important questions that we would like to be able to answer.

1) What were the causes, both immediate and mediate?
2) What can be done to correct the problem?
3) How can this be prevented from happening again?

There is a whole litany of reasons being thrown around as answers to the first question. Depending on who you listen to, you can hear that the problem is a result of credit derivatives, securitization, Wall Street bonus payments, quantitative financial engineering, failure of government regulators, government sponsored mortgage agencies, credit rating agencies, consolidation of the financial industry, American consumer demand, unscrupulous lenders, the abandonment of the gold standard, the Federal Reserve’s easy monetary policy, and many more. In reality, all of those things did contribute in some way. You can take any item from that list and use it as a window into understanding the phenomenon, but you will quickly encounter more and more of these items until you realize that a holistic approach is needed. That is to say, you can not understand or address the problem by focusing on a single aspect of it because it is too big and too complex. You need to understand the dynamic relationship between all of these things.

Let me start by emphasizing that this crisis was in no way caused by a conspiracy or anything resembling one. Of this I am certain. Rather, it is the outcome of many different people and institutions individually taking logical and reasonable steps to look out for themselves. This is a complex structure which emerged from many smaller, random actions.

I do not want the scope of this discussion to grow too large, so I will start at a time around 2001 and leave the analysis of how we got to that particular point in history for elsewhere. This time is significant for a few reasons. For one, the US economy had just been dealt a major setback by the collapse of the technology sector, along with the failure of a few large firms due to massive fraud, such as Enron and Worldcom. In addition, the terrorist attack of Sept. 11 represented another setback, at least psychologically. Finally, this period coincides approximately with the introduction of credit default swaps (CDS), an important type of credit derivative.

Interest rates in the US, which by historical standard were already fairly low, were cut even lower by the Fed in response the threat of recession. This is the standard action that you would expect from the central bank when they believe that the threat of recession outweighs the threat of inflation. In this case, it worked, as the US economy resumed growing and stock prices recovered and began to recoup some of the losses from the tech crash. In fact, it may have worked too well.

Under normal conditions, the Fed sets the price of money. Low interest rates mean that money is cheap is the sense that you can borrow it quite easily. This creates an especially profitably environment for banks. Banks make money by borrowing cheaply and lending at a higher rate. This is true for almost all aspects of the banking business, although the complexity of the borrowing and lending mechanisms can get very intricate for some of the activities of big investment banks. In order to increase profits, banks can increase the spread (the difference between what they receive on the loans they have made vs. what they pay on funds they have borrowed), and they can increase the volume of business or the size of the transactions. The first of these is difficult, as the competitive nature of the industry tends to compress the spreads of financial products rapidly, so the majority of the banks’ strategy focuses on doing more and larger transactions.

The easy money environment created a situation where there was a huge demand to borrow money (from businesses, home buyers, consumers.) However, banks can not indefinitely lend money because they are required to maintain capital reserves equal to a percentage of the amount they lend. For instance, when a bank loans $1,000,000 it creates both an asset and a liability. The asset is the obligation of the borrower to repay $1,000,000 plus interest. The liability is created when the bank itself borrows $1,000,000 to give to the borrower (through customer deposit accounts or various other sources), on which it owes $1,000,000 plus interest (a lower interest rate than what the bank charges the borrower.) Because the interest rates are different, the value of the asset is greater than the value of the liability, and the bank has made a profit, at least on paper. However, there is a risk that the borrower will not repay the loan, and therefore bank may have to reserve $100,000 against this risk.

The reserve requirement is the mechanism which rations credit in the economy, forcing banks to make the loans that are most likely to realize an actual profit, given the risks involved. Those capital reserve requirements in turn are determined by the Federal Reserve Board based on international banking standards commonly referred to as the Basel Accord. The main idea behind the Basel Accord is that banks should reserve more capital for loans with greater risk. The riskiness of a particular debt is usually determined by one or more of what are known as rating agencies. The two main agencies used to determine capital requirements are Moody’s and Standard & Poor’s, with Fitch as an occasional third. It is important to note that these are private companies, not government entities, about which we will say more later.

These circumstances created a situation where the demand for credit exceeded the supply. Traditional economic theory would lead you to expect that the price of credit should go up until demand and supply are equal. However, there are some important ways in which credit differs from a normal economic good. One of these is the fact the Federal Reserve, to a large degree, can set the price through monetary policy and certain other measures. In addition to the Fed, there are other government policies that directly or indirectly influence prices for different segments of the credit market (for example, the various in ways in which the government subsidizes borrowing money to buy a home.)

Thus, private lending institutions were not in a position to raise the price, at least not by enough to equate supply and demand. Instead, they did what economics and logic would suggest; they found ways to increase the supply.

This step is where it becomes important to understand two important innovations in financial technology. The first of these had already been around in its modern form for a few decades, but found a whole world of new applications when combined with the second innovation.

The first is called securitization. Activities resembling modern securitization have been around at least since the 18th century, and quite possibly earlier, but the thing that we think of today when we say this word had its origins in the early 1970’s in the mortgage markets. In its most basic form, securitization involves taking a pool of assets, such as mortgage loans, and packaging them together into a financial security, usually for the purpose of selling shares in the resulting asset-backed security. The first modern securitized products were mortgage loans backed by the government sponsored entity known as Ginnie Mae. The resulting security can be as simple as a pass-through, which as the name suggests allows all cash flows on the underlying assets to pass through to the owners of the security. Or it can be made arbitrarily complex by creating multiple securities of different classes based on the underlying asset pool, with rules regarding which classes get paid how much and when based on the performance of the underlying assets.

The second is called a credit default swap (CDS), a type of credit derivatives. Credit derivatives, in various forms, have also been around for a long time. However, the use of a swap form was a new innovation that happened around this time, created by JP Morgan. Swaps had existed in the interest rate and foreign exchange markets since at least the 1980’s, but had never been applied to credit. Simply put, a swap is an agreement between to parties to exchange payments at certain dates based on the value of certain underlying quantities, usually interest rates or currency rates, at least until the invention of the CDS. For example, in a very simple swap between A and B, A agrees to pay 5% of $1,000,000 in 3 months to B, while B agrees to pay the interest rate on the 10 year US treasury bond to A on that same date. In 3 months, if the10 year treasury is below 5%, A will make a net payment to B, otherwise B will make a net payment to A. This is a simple interest rate swap. The value of $1,000,000 which is used to calculate the amount of the payments is called the notional, and it is important that the notional amount never actually changes hands.

The CDS is a swap that depends on whether or not a particular debt is successfully repaid. In CDS, the 2 parties are called the buyer and seller of protection. The protection seller typically receives a regular payment which is a percentage of the notional, until either the contract expires, or the underlying debt defaults. If there is a default, then the protection buyer delivers the defaulted bond (typically worth only a fraction of its initial face value), and the protection seller pays the full initial face value. In practice, these contracts are often settled by the seller paying a cash amount equal to the difference between the full value and the value of the defaulted bond, without an exchange of the actual defaulted bond ever taking place.

In order to understand how these 2 innovations came to interact and how they are important to the current crisis, it is necessary to understand one more financial instrument. It is called a Collateralized Debt Obligation (CDO), and it is a direct outgrowth of the securitization process. The CDO takes a pool of bonds (or other debt instruments, loans, etc), packages them together, and creates several classes of securities. These classes are usually referred to as tranches, and while it practice there are often 5 or more tranches in a CDO it is only important to distinguish between 2 types right now.

The first type is the junior tranche, sometimes called the equity for reasons that will soon be apparent. The equity tranche absorbs all of the losses on the underlying asset pool until is completely exhausted. If the underlying pool has 100 bonds in it, and the equity is defined as the first 10% of losses, then each bond that defaults will erode 10% of the value of the equity (this is a special case where there is no recovery on the underlying bonds.) After 10% of the underlying bonds have defaulted (10 bonds), the equity is entirely exhausted and has no more value. The equity tranche bears the great majority of the risk in the underlying pool, and therefore generates high returns for investors if the underlying assets do not default. Returns of 20% or more annually were typical of CDO equity in the mid-2000’s.

The second type is the senior tranche. If the equity absorbs the first 10% of losses, then the senior tranche absorbs all losses in excess of 10%. If 10 out of 100 bonds default, the senior tranche will not have lost any value. Typically, an assessment of the risk of the underlying pool would determine the probabilities of losses exceeding a certain amount, and the division between equity and senior would be chosen so that the senior tranche would be insulated from any loss with a fairly high probability.

In this manner, a pool of risky debt, for example BBB rated corporate bonds or subprime mortgages, could be changed into 2 securities: an equity piece with the majority of the risk, and a senior piece with very little risk.

The determination of exactly how to split the pool was heavily dependent on the rating agencies. The goal was to re-classify a large senior tranche as a AAA rated security (the highest credit rating, given to only a small handful of corporate borrowers and some of the most stable sovereign borrowers such as the US government.) In fact, rating agency methodology played a large role in the evolution of these products. Remember that the determination of the reserve capital is based on the credit rating from the rating agencies, and the amount of reserve capital directly controls how much lending a bank can engage in. If a bank can take a pool of risky debt and turn 10% of it into equity and the other 90% into virtually risk-free debt (at least in the view of the rating agencies), it is then in a position where, if it can find someone to take the risk of the small equity piece, the bank will get rid of nearly all of the risk and therefore nearly all of the reserve requirement of the whole asset pool.

Both CDOs and CDS present interesting mathematical and computational challenges. In the case of the CDO, complicated assumptions about the risk of the underlying assets, and how those assets may be correlated with each other, enter into the valuation and risk-analytics necessary to figure out how much they are worth and how they should be hedged. In the case of the CDS, the random and unknown timing of the default event makes them more difficult to value and hedge than interest rate or FX swaps. For these reasons, many mathematically inclined financiers, the so-called “quants”, were brought to bear on this market. Indeed, the number of quants hired by Wall Street increased dramatically throughout this period.

Finally, the ultimate risk of the all these loans had to end up somewhere. Asset managers investing retirement savings, pension plans, and insurance premiums have historically been the biggest investors in the market for credit risk, although these parties can also include sovereign governments or other entities. These asset managers need to meet future liabilities (pension payments, insurance claims, etc), so they take the cash they generate today, invest it in fixed income securities, and create a future income stream to meet liabilities. They typically must make a decision about how to allocate their investments between government bonds (referred to as interest rate products, since they bear no default risk), and credit products including mortgages, corporate debt, emerging market debt, and other types of risky debt.

When interest rates are low, it is impossible for the asset managers to meet their liabilities based solely on the returns they can earn from risk-free government bonds. The long period of very low rates, dating back to the Greenspan Fed’s response to the collapse of the internet bubble, caused asset managers to allocate increasingly more of their assets to the risky credit products in order to earn a spread over the risk free rate.
However, as the lenders competed to increase market share and overall supply of credit, they actually drove the credit spreads tighter at the very time when they should have been going wider. That meant that even the rate paid on a risky bond was no longer enough to meet the targets of the asset managers, as the combination of low rates and compressed spreads kept overall yields at such a depressed level.

One way to increase yield is to invest in riskier assets, called going down in quality. However, most pension funds and insurance companies have institutional limits on the minimum quality of the assets they can hold. These limits are determined by the credit ratings of the assets. As asset managers were hitting their limits for credit risk, and still looking for extra yield, Wall Street began providing structured credit products, most prominently variations on the CDO. It was now possible to take a pool of very risky assets, create an asset-backed security, and split off a senior tranche which would attain a very high (non-risky) credit rating from the rating agencies. The spread on this tranche would often be multiple times the spread on an equivalently rated non-structured bond. For example, AAA spreads had been driven down as low as 15-20 basis points (hundredths of a percent) on bonds, while structured notes could pay 60-70 basis points.

Further innovations allowed these spreads to be pushed even wider relative to the equivalent non-structured investments. Synthetic CDOs were created which were not backed by any asset pools at all, but instead by pools of derivatives, typically credit default swaps. Thus a structured derivative based on the value of a pool of other derivatives could be tranched, and a AAA rated tranche could be created. Investors could buy this tranche as a derivative transaction (a swap), or else it could be transformed into something that looked like a bond but still had the same characteristics as the derivative, usually done to satisfy regulatory requirements. The spreads on the synthetic tranches could be as much double the spreads on the tranches backed by actual bonds, which were already several times wider than the equivalently rated bonds themselves.

How was this possible? There were a lot of tricks and tactics, but the main ingredient was leverage. Investments in tranched credit portfolios can almost always be reduced to simple leveraged exposure to the underlying credit risk. An investor can choose between buying a$1 million AAA rated bond, or borrowing $2 million secured by the $1 million cash and buying a $2 million AAA bond. As long as the bond performs (and for AAA rated bonds there is virtually no chance that will fail to make all promised payments), the second scenario provides a far greater return on assets. Most asset managers are highly restricted in their ability to take outright leverage. However, structured credit derivatives provide implicit leverage through the nature of the structure. The asset manager may not be allowed to borrow $2 against $1 of assets , but if he can buy $1 of AAA rated synthetic CDO paper he has essentially done the same thing without calling it by name. In practice, the multiples gained by the implicit leverage were often much higher than 2:1. When a company levers up its balance sheet by borrowing more money, it will typically receive a lower credit rating to reflect the increased risk. Certain types of credit derivatives were able maintain very safe ratings even while taking on enormous implicit leverage.

These types of financial instruments represent only a sample of the full variety of structures that were created and sold during the last decade. Once securitized products were combined with CDS derivatives the complexity and amount of the resulting products could increase as quickly as the documentation could be drafted. A typical structured credit derivatives trading desk during this time might include a few traders who constantly survey the market to gauge changes in supply and demand; a couple of individuals with a background in law or one of the ratings agencies, sometimes called structurers, who try to find new ways to exploit regulatory loopholes; a few quants, who are hybrid mathematicians-computer programmers, and who create and maintain the software which calculates the value and risk of the derivatives; and of course a battalion of salespeople to place this product with investors.

So, to step back for a minute, let’s take a look at the economics of this market from Wall Street’s perspective. Clients (the asset managers) wanted rated bonds that pay a wider spread than available bonds, due to the compression of both interest rates and credit spreads. In order to meet this demand, dealers created structured credit and credit derivative investments that maintained high ratings while taking on large amounts of leverage in order to dramatically increase spreads. One consequence was that the dealers needed a constant supply of the underlying product (the actual loans) which the actual cost of credit in the real economy very low (reflected in low rates for mortgages, consumer borrowing, and corporate borrowing.)

The dealers could then sell the rated paper (sometimes called the mezzanine) to asset managers, and pay them a spread that was much wider than the equivalent bond, but still only a fraction of the spread based on the actual risk of the investment. The residual spread was retained by the dealers as profit. For example, a AAA structured investment that had an actual leverage of 5x the underlying asset pool might pay 2x ot the investor and retain 3x for the dealer. Some of this retained value was needed to cover the costs of originating and hedging the risk over the life of the transaction, and the rest of it was booked as profit. Asset managers were happy with this because they measured their risk in terms of ratings, which indicated the investments were very safe.

The ability to sell the middle tranche of the asset pool at a large profit drove the growth of the market. However, dealers were then left with unrated, or equity, tranche as well as the risk beyond the AAA rating, sometime referred to as a super-senior tranche. For example, a typical pool of assets might have rated tranches from 5% to 15% of portfolio losses, with the first 5% as unrated equity and the final 85% as super-senior.

The equity, which contained the majority of the actual risk, was the focus of much risk management. Many hedge funds found a strategy whereby they could take the equity risk of these deals in return for a cut of the profits the dealers made from selling the mezzanine. The dealers effectively outsourced the risk management to the hedge funds, allowing them to do more transactions.

The super-senior was perceived as having no actual risk. Dealers usually engaged in some combination of keeping these tranches on their own balance sheets (which was free, since they were riskless), or selling them to a new class of investors that essentially did not exist. The super senior tranche paid a much lower spread than what was justified by the market implied risk, but the ratings based risk was zero so any spread above zero was considered “free money” on this basis. A class of investors formed who agreed to take the super senior, but would put up no collateral, in return for the tiny spread. They made money by doing this in huge size. In the event that the super-senior was actually impacted by the performance of the underlying securities, these investors would not have even a fraction of the capital required to pay for the synthetic bonds. Trading with these types of quasi-fictional counterparties was not much different from simply retaining the risk on the balance sheet for the dealers.

So the situation arose where the dealers were demanding massive amounts of underlying loans, thereby keeping credit spreads tight, in order to engage in profitable mezzanine transactions with hidden leverage. Rating agencies continued to affirm that these deals had very low risk. They paid some of these profits to hedge funds in order to get rid of the majority of the resulting risk, and kept the very low-probability risk, for which they were underpaid by construction, on their own balance sheets.

The thing about financial leverage is that, when times are good, it becomes necessary to engage in more and more in order to produce the required profits. It is a self-reinforcing cycle that higher leverage creates more demand for the underlying assets (loans in this case), which in turn pushes the value of those assets higher. In the bond world, higher asset value means lower spreads, and therefore it takes even more leverage to achieve the same ultimate return on investment.

This situation also accelerated a trend that was already happening whereby the originators of the loans kept none of the actual risk. Lending institutions which used to have a direct interest in assessing the credit risk of the borrower now had an interest in making as many loans as possible, collecting the fees, and selling these loans to Wall Street banks who needed them as the basis for structured, securitized products. The natural limit on the amount and riskiness of the loans being made, i.e. the willingness of the lending institution to take on the credit risk, was obviated.

For consumers, this was a great situation. As the value of assets rose, namely real estate values, owning a home became a money pump. In an easy credit environment, people could borrow the entire value of the asset in order to purchase it, and then continue to borrow against the asset as it appreciated in value. The resulting cash flows fueled consumer demand, and were the primary engine of growth in the economy.

As long as asset prices continued to appreciate, the cycle continued. More loans were made under increasingly lax credit standards; these loans became the basis for more complex and highly leveraged structured derivatives; Investors continued to hold more of the resulting securities because they were rated as extremely creditworthy by the rating agencies; and Wall Street firms continued to generate huge profits while stockpiling massive, hidden risks on their balance sheets.

When the music stops, however, and asset prices stop going up, it become like a game of musical chairs in which all of the chairs are gone. The more leverage there is in a system, whether it is a single household or corporation, or an entire government or global economy, the less that system can tolerate deviations from the expected rate of asset growth. We will probably never know exactly why mortgage backed assets started to worsen exactly when they did. It may be related to the increase in the price of energy and commodities which happened at around the same time, or to some other external shock. The performance of a highly securitized product like a mortgage pool is based on the individual decisions of many thousands of people.

However, once one sector began to show weakness (the sub-prime mortgages), the problems spread quickly and violently. The floor fell out of the housing market in many areas, leading to foreclosures and a vicious spiral of declining prices. The banks which had been creating all of these structured assets, remember, had also been directly and indirectly retaining much of the risk. When the underlying mortgages stopped performing, the value of these assets on the bank balance sheets declined precipitously. The game had always been to undercompensate the holder of the most senior risk in order to profit off the junior and mezzanine risk, but the banks themselves had forgotten their own rules and were left retaining much of the senior risk.

The main impact to the financial economy was a sudden and drastic credit freeze. Most banks avoided an official re-pricing of their assets by continuing to mark them based on outdated assumptions that were fed into software models. They refused to actually transact to get the assets off the balance sheet, because to do so would require crystallizing the magnitude of the losses. Many banks, especially some of the largest ones, would be insolvent if forced to write down assets to the levels where they could actually transact.

The real economy, which is the term for the part of the economy that makes non-financial goods and services, was now hit by two damaging effects. First, the rapid disappearance of credit meant that most companies could no longer finance their operations. All companies must pay for operational costs such as materials, labor, rent, power, etc. They can pay out of the cash flow generated from operations, but this is inefficient and usually only done by the smallest mom and pop operations. Cash flow from operations can be chunky, while operating costs tend to be steadier. The solution is that most companies rely on frequently rolling short-term credit lines. These credit lines disappeared because banks were unwilling to lend, which in turn jeopardized the ability of many companies to meet basic expenses.

At the same time, the consumer demand engine of growth had just been shut off. American households have had a savings rate around zero to slightly negative for about a decade. Because of the increase in household wealth caused by the stock market, and then home values, most people did not feel the need to save any of their income, and in fact many borrowed against the value of their home to spend more than their income. As both of those sources of wealth were vaporized, households quickly began saving again. The additional fear caused by layoffs and pay cuts only exacerbated the problem.

The response by the government contains a number of key elements.

The fed aggressively cut rates and purchased securities in the market in order to keep rates low. Low rates generally encourage economic growth, and current rates are near zero. This time, however, banks are not willing to lend and so the benefits of the low rates are accruing mostly to the profits of the financial institutions. In fact, profits have been near record levels for most of the banks still operating in 2009. The level of rates allows banks to borrow money essentially for free, which they then can either lend out to private enterprise to earn a profit, or they can simply lend it back to the government to take advantage of the steepness of the yield curve. Most are choosing the latter option, as it does not put any of their capital at risk.

For example, a bank can borrow money at 0% (through its deposit base, for example), and lend money back to the government via purchasing 10 year bonds, at around 4%. Since government bonds are considered risk-free under the capital regulatory requirements, the bank has not used any balance sheet but it has generated a positive cash flow. The alternative of lending money to corporation at 10% requires the bank to reserve capital against the risky loan. An undercapitalized bank can not afford to commit more balance sheet to a risky loan.

The Treasury has engaged in a series of bailouts, including the TARP program, several payments to insurer AIG, and various payments to the auto manufacturers. These bailouts are direct cash payments from the government to specific firms. They are a logical solution to a liquidity problem. If a company has positive value and profitable operations, but needs money immediately to pay for current operations, it is said to need liquidity. When it can’t turn to short term financing, it has a liquidity problem. If a company is not profitable, then adding liquidity can only postpone its inevitable bankruptcy.

In addition, there have been some modifications to the accounting rules that allow companies, mainly financial companies, much greater latitude in assigning value to assets on their books. In particular, it allows banks to continue holding complicated asset-backed and structured products at values that they think are fair, as opposed to values where they could actually transact. The PPIF was an attempt to determine accurate valuation for some of these securities by allowing private investors to bid on them with government-backed financing, but it appears not to have gotten off the ground. Allowing firms to mark their assets at higher values forestalls recognizing that they are insolvent, and buys time during which they can hope for the market to recover.

A healthy financial sector is necessary for the function and growth of the economy as a whole. However, the financial sector is very powerful because it controls the supply and production of credit, and therefore is subject to regulation. It is my opinion that the quickest way to get this sector healthy is to allow banks and financial institutions to fail. Continuing injections of public money into these institutions only postpones this result while squandering resources. The banks main problem is not a liquidity constraint, it is that many of them are insolvent.

The that end, it is also counterproductive to allow assets to be valued wherever banks want to value them. Banks need to write down assets to reflect their market value, not their wished-for value, even though this will result in insolvency for many of them. Institutions which are insolvent can then be nationalized, run by an institution such as the FDIC, and re-privatized when possible. This practice is followed for most small and medium sized institutions, but for the enormous multi-national institutions there are dangerous exceptions being made.

In terms of regulation, I think most of the new regulation being proposed is of minimal value and in fact is probably harmful. I see no reason to believe that institutions like the SEC are or could be competent to identify systemic problems in the global financial system in time to act upon them. Increased regulation will definitely have the effect of slowing down financial innovation, which would be unfortunate, and of creating more incentives to circumvent the regulation, which would enrich a small group of layers, structurers, and traders at the expense of distorting the economy. The government agencies tasked with regulating the markets have proven themselves to lack the competence, foresight, and ethics to achieve their goals.

The rating agencies were guilty of providing false signals as to the safety of the credit instruments they were rating. This is hardly surprising considering that their main source revenue came from the Wall Street firms originating these instruments. Most trading desks employed at least one person with experience working at either S&P or Moody’s. Working together with quant programmers and traders, these teams could easily exploit the loopholes in the rating agency assumptions to pack as much leverage as possible into a given credit rating. The rating agencies were well aware of this and even facilitated it by making their software available to all clients.

The main problem with the rating agencies lies in setting investor guidelines based on such a biased source of information. While the credit rating can provide some very gross information about the underlying risk, investors must be responsible for assessing risk independently. If an investment manager does not have the expertise to evaluate the risk of a complex transaction, then he should probably avoid that transaction.

Among all the other calls for regulation and systematic overhaul have come demands to regulate the pay of those on Wall Street who created and marketed the suite of structured products and derivatives. I believe that the incentive structure at these firms is very badly skewed, and that it did play a major role in the problems we face today. However, I do not believe that regulation of pay will solve the problem.

First of all, there is the problem of enforceability. Regulating certain entities creates an incentive to take the most profitable business to unregulated entities. This is the trend which fueled most of the growth in hedge funds during the last 10 years. In addition, it is unlikely that legal regulation can keep up with financial innovation, so there will likely exist structures which allows people to be paid large amounts of money while satisfying the demands of whatever regulation is in place.

Second, is the owners of these firms who should have the greatest interest in changing the incentive structure of those who work for them. When an employee can take out millions in cash based on the performance during the last 12 months, even while leaving highly risky assets on books that extend out for 5, 10, or more years, the employee is enriching himself at the expense of the firm.

For example, a trader on the credit derivatives desk of a large bank would typically get paid based on the profitability of the desk as the dominant factor. This payout happens every 12 months, based on the performance of the last 12 months. Many of the transactions involve contracts that extend far longer than 12 months, and whose value is based on feeding noisy and imprecise market data, along with unobservable parameters, into software models. Thus, the trader may find that 10 year contracts are highly profitable according to the bank’s models. The fact that the actual value of these contracts is uncertain is only partially accounted for, and if it turns out that the value is far less than expected, or even negative, that is not likely to manifest for several years. In the meantime, the trader has already been paid and cashed the check.

Every bank contains a risk group whose role is, in theory, to challenge the valuation of the trading and sales groups and to protect the interest of firm from the potential gaming of the employees. If the value of a transaction is very uncertain, the risk group should force the trading group to reserve that value, so they can not recognize the profit and be paid for it until it becomes more certain.

In practice, the risk group is very weak relative tot he business unit that produces the profit. Trading and sales make money and therefore have far more influence. They can effectively shut down the risk group in almost all cases. The pay and prestige of the trading desk are far greater than the risk group. The most talented individuals in the risk group typical move to the trading desk, where they can outmaneuver their peers who are left behind.

Currently, all the major banks and Wall Street firms are public companies. Perhaps it is the diffusion of ownership across millions of stockholders which creates the problem. A more concentrated ownership structure, such as the partnership structure that prevailed for centuries in the US and the UK, may be more conducive to aligning employee incentives with those of the firm. This is especially true in industries where there is often a lot of uncertainty around the actual value of what is produced, and where the employees expect to receive a substantial portion of the value they create. Tax and regulatory policy which encourage concentrated, partnership structures over diffuse public ownership may be the best way to align Wall Street pay with the interests of the firm and the economy as a whole.

---

Written by & Posted on April 24th, 2010

---