The risk of price manipulation of securities is likely the top banking risk that is least understood by the layman. Indeed, at first blush, the sin of deliberately mispricing a security does not appear the most egregious of financial sins. Yet, as we have seen with the LIBOR scandal, it can lead to significant market distortions, price cheating, and significant penalties from regulators. Understanding the reason why that is so and what can be done to mitigate the risk is one of the most urgent issues for risk managers on Wall Street today.

What catches the headlines is generally the bad actions of a few. However, there are many, many people involved in complex, difficult valuation work that do the right thing every day. Someone wandering through a typical fund accounting operation, for instance, would perhaps be struck by the level of focus and intense effort expended on the activity. Controllers spend a great deal of the day pouring over the valuation of securities in each portfolio. However, any security ultimately depends on there being a market to purchase it. What happens when no one wants to buy the security in question? A precipitous drop in prices can occur. This is what happened to credit securities and subprime securities in the summer of 2007.¹

Many traders and many firms at that time were depending on the buoyant market in mortgage securities to drive profits and revenue. Instead, these prices suddenly dropped due to change in housing market and associated investment products, and the drop in prices affected the salaries of many. In a market for equities that has typical levels of transparency, the price cannot be hidden or run away from—it is what it is. In the case of mortgage‐based credit funds, however, prices depend on the quotes received from the brokers who deal in these securities. These brokers are able to provide prices because they are active in buying and selling these types of securities. When prices started to drop early in the second quarter of 2007, there was much disbelief and skepticism. In a very short time, however, based on the quotes coming back from the brokers, certain securities lost almost all of their value. As funds and banks completed the painstakingly difficult and concentrated work to finalize the impact of the price changes, the loss of values, for the vast majority who held long positions, was an extremely difficult message to publish and not an easy position to be in. In just one example, the valuation of the Bear Stearns High Grade Structured Credit Fund was destroyed almost overnight.² The difficulty in communicating what had happened to management, to investors, to the outside world was immense. Not just because the results were so unpalatable for investors but also because they were met by a world with no understanding of the cause. While investors, everyone, can hear every day, every minute, of the gyrations of the stock market, there is no such way for them to gauge or understand the valuations prevailing in the market for certain complex or exotic securities. That is still the case, and that is part of why this risk is still significant for Wall Street. In a world where certain types of securities do not have an easily agreed upon value, there is temptation to manipulate valuations.

Price Manipulation

There are instances where traders are willing and able, on occasion, to manipulate the price of a traded security. Such instances are able to happen when securities are complex, hard to value, and are not traded on a public exchange. In general, of course, trading in marketable securities happens when people can agree on a price to execute a transaction. Most of the time, that price is broadly determined by an exchange and can be tested by analysts and traders on the basis of easily observable data in the market. Transparency is thus facilitated by an exchange and pricing data available to all market participants. Certain securities, however, are valued on the basis of price information that is not so transparent to others. Pricing may, in fact, be dependent on expertise or market data available to a restricted number of persons. Transactions in such securities may also occur not on an exchange but on the basis of a bilateral agreement between two parties. Exotic securities such as structured notes, collateralized debt obligations (CDOs), and credit default swaps (CDS) have been typical of such transactions up to now (though this is now changing for some type of swaps, thanks to the Dodd‐Frank legislation). Banks also provide valuation expertise and pricing information that becomes an input into the valuation of securities and assets. The interest rates at which banks lend to one another and use as benchmarks for other transactions are an example of such valuation services. LIBOR (London Inter‐Bank Offered Rate) is the most well‐known of these, but there are many others.

It turns out that banks run several risks in this area. First, in the case of transactions in securities that are hard to value, banks run a risk that is hard to measure when they lack an independent and authoritative valuation source for a given security. (Of course, this also has another side to it—an opportunity to realize excess profits.) In such rare cases, traders may be able to, if they are tempted by opportunities to improve their earnings for the year, manipulate the price of the security they hold in their favor. Second, in the case of valuation services, banks are at risk of introducing distortions into the marketplace if they fail to provide objective valuation data. Let's look at three major examples of the several past years.

Earnings Restatements

In early 2008, Credit Suisse restated its earnings from Q3 2007 because of an apparent miscalculation. Restatements like this occur rarely, and it happened in this case in large part due to the deliberately inaccurate valuation of a credit derivative portfolio. Three traders, Salmaan Siddiqui, David Higgs, and Kareem Serageldin, later pled guilty in a Manhattan courtroom to charges of deliberately marking their credit derivatives portfolio at inflated prices to hide hundreds of millions of dollars in losses.³ How could something like this happen?

Owing to the lack of publicly available pricing data, and in common with general market practice, the Credit Suisse credit portfolio was valued based on the traders obtaining bid and offer quotes from other dealers in the market. In reality, however, according to the criminal complaint, the traders themselves priced the portfolio and the quotes they were using were prices from earlier in 2007, before the beginning of the housing market collapse. The difference in the valuation of the securities once more appropriate quotes were applied was $1.3 billion. In other words, something like a market crash had taken place since the prior quarter, but the portfolio was being valued by these traders as if the market was still buoyant. That way their positions looked good. The interesting thing, and also rather worrying, is that apparently none of the people mandated to provide oversight of the valuation process—in particular, financial controllers, risk managers, and trading supervisors—knew enough to question the values those traders had assigned to the portfolio, as those values were accepted as part of the original financial books for the quarter.

LIBOR

Several banks were in the headlines in 2012 for indiscretions in the setting of interest rates, most notably the LIBOR rate.⁴ The charge was that these banks formally admitted that “the manipulation of the (rate) submissions affected the fixed rates on some occasions.” The benefit to these banks was, first, to make their businesses appear stronger than they actually were. A lower interest rate implies a perception on the part of the counterparty of a stronger balance sheet, helping to prop up the bank in question through the financial crisis. The second motivation was to trade on small differences in interest rates based on the certain knowledge of what interest rates would be.

Foreign Exchange

In 2015, it was announced that six global banks were to pay more than $5.6 billion to settle allegations that they rigged foreign exchange markets, in a scandal the FBI said involved criminality “on a massive scale.”⁵ Announcing the settlement, the US Department of Justice said that between December 2007 and January 2013, traders at Citigroup, JPMorgan Chase, Barclays, and the Royal Bank of Scotland, who described themselves as “The Cartel,” used an exclusive chat room and coded language to manipulate benchmark exchange rates “in an effort to increase their profits.” By agreeing to place orders at a certain time or sharing confidential information, it was possible to move prices more sharply versus working alone where the impact would be smaller. Such price changes close to the time of the pegging of the rate daily could result in advantageous positions, for example, versus client orders currently being worked by traders. That could result in traders making more profits and clients being hurt (see Table 5‐1).

Collusion

What all of these cases have in common is a pricing mechanism that is vulnerable to exploitation by manual intervention and collusion between those with a self‐interest in doing so. Collusion was active during the FX incident, with traders speaking to each other on the phone or on Internet chat rooms. It can also be implicit, where traders don't need to speak to each other but are still aware of what other people in the market are planning to do. The revelation that traders colluded to move around currency exchange rates was particularly embarrassing for the banks because it occurred after they had paid billions of dollars to settle claims that their traders had tried to rig interbank lending rates. It also raised questions as to whether the industry learned any lessons from the LIBOR or the credit securities scandals.

Some of the key risk indicators and controls included in Figure 5.1 include the following:

• Traders are able to use two different systems to trade fixed‐income securities, allowing them to arbitrage executions between the two.
• Traders price their securities using nonapproved pricing models.
• Security price levels do not change even though markets have.
• Unusual trading activity occurs just before price fixing (usually at market close).
• Trader has access to control and edit price models.

Although there are many risk indicators, the issue continues to be the opportunity for deliberate price manipulation owing to the obscurity of the valuation drivers. Many of the incidents that have been subject to recent settlements between banks and regulators date back to 2008 or even earlier. Controls have been strengthened since that time, in part due to pressure from regulators to make such improvements. Key controls in Model Risk and Control center on price models and independent price verification. Banks employ large teams of administrators responsible for ensuring that each position on the books is associated with a certified valuation model and authoritative data source. In addition, price testing is generally conducted on each position on a monthly or quarterly basis by an independent valuation unit.

Emphasis on Surveillance

Banks have long had trade surveillance teams in place to cover the equity trading markets, with the primary objective of uncovering evidence of market manipulation through insider trading, front running, and so on. The lesson learned from the FX and LIBOR scandals is that, first, the FX and LIBOR markets can be manipulated, and second, surveillance is required both of traders and their communications in these markets. Indeed, since these scandals took place, regulators have been requiring via consent orders that banks extend and modernize their trade surveillance capabilities to these additional markets. We will discuss in Chapter 21 the opportunities provided by modern computing tools and techniques, particularly in the AI and cognitive space, to strengthen surveillance capabilities.

Despite efforts to strengthen controls, the increased rigor of Model Risk Control being a central element, banks still have exposure to price manipulation risk owing to continuing reliance on manual processes and human intervention for certain hard‐to‐value securities. Risk managers, therefore, need to continue to focus on improving controls in this space, in particular, to ensure that independent valuation control groups are able to attract high‐caliber valuation experts, that traders are not able to manage the pricing process directly, that valuation models are certified and consistently followed, and that manual processes are kept to a minimum.

In the future, the exposure that banks have to this risk should be reduced by the drive toward a more transparent marketplace. We look at this in more detail in Chapter 17.