“It is when our clients are the most fragile that we make the most money.” Part two in a series of columns by anonymous sources that reveal how industries work from within.
As a former London employee of a major investment bank, I am often puzzled by the tone that top managers of investment banks use when speaking to the public. There is indeed a striking gap between the official communication and the internal behaviors I have observed (and taken part in). To me, banks are experts at exploiting asymmetries of information. Furthermore, they often amplify this asymmetry themselves by complexifying the products they offer or by disclosing only fractions of the information they have.
Of course, investment banks’ clients are the principal target of this type of strategy. While banks typically claim as their main value that their clients’ interests always come first, the reality is usually quite different. Therefore, banks will routinely increase the complexity of a transaction to make it difficult for the client to understand where the bank makes money or to make it difficult to compare to transactions proposed by other banks. A major step in this direction by my former employer was to develop proprietary financial indices. Proprietary indices are financial indices that rely on a pre-determined investment strategy implemented by the bank. For instance, the bank will renew the same position in a derivative every three months, and the index will track the performance of this strategy over time. While the marketing angle of these products was that the client benefited from the bank’s expertise, the primary purpose of the whole product line was to shield the bank from competition: no one else could offer the same transaction at a lower price, simply because no one else could offer this transaction. This also put the bank in a position of a monopoly when a client wanted to exit one transaction early. Another classic strategy was to focus the client’s attention on a loss leader product—say, a bond issuance—where fees are low and disclosed, while cross-selling a derivative—for instance, to neutralize the foreign exchange exposure, which embeds a large and undisclosed markup.
Making one aspect of a transaction salient was another key phenomenon. For instance, we once designed a transaction that was a mix of interest rate and foreign exchange exposure. The risk came from the foreign exchange exposure, but we could argue that the product was an interest rate product and that foreign exchange offered diversification. Another pillar of our marketing strategies was to design products that would look favorable in a back-testing test—i.e., when looking at how it would have performed historically. However, the current market conditions were typically pricing the opposite evolution compared with the historical one, thereby generating value for a transaction betting that the historical level of a given financial asset will prevail. The favorable back-testing helped us make a case for how attractive the product was, and it was also required by the investment product regulation. For instance, we would market the product that offered positive payments when the yield curve was steep, as it had been the case historically, even though the forward interest rates at the time of transaction were pricing a flattening of the yield curve, if not an inversion. The yield curve did invert shortly after.
It is important, however, to note that most marketing strategies are comparable to the ones used by consumer goods companies, such as hotel or car rental pricing strategies. The term “structured,” which was ubiquitous in titles and product lines before the crisis, could easily have been replaced by “rent-extracting.” For instance, Structured Capital Markets was the division designing tax-optimizing transactions—i.e., structures between banks with no economic substance but allowing the banks to get tax credits by arbitraging differences in tax regimes between jurisdictions. Structured derivatives led to high profitability to the bank, as the more the product inflated the client’s expectations, the more the bank could profit. My boss used to refer to it as “monetizing indifference zones,” but I believe that obfuscation would also be an appropriate qualification for the rationale of these transactions.
When thinking about answering clients’ needs, plenty of examples illustrate how this was only a secondary objective for the bank. One of my colleagues seriously proposed that a corporate client should borrow through a structured product, even though the client did not need any funding. The reply from my colleague when the client raised the point was disarming: it could invest the proceeds with us in another product.
Another classic strategy was to take advantage of our clients’ weaknesses. Thus, during an annual internal keynote in 2007, the global head of corporate finance said bluntly to our division that “it is when our clients are the most fragile that we make the most money.” It worked like this: if a client desperately needed something from us—say, some funding—you would provide it only if the client agreed to an additional transaction that he did not necessarily have an appetite for but that was much more profitable. A good example was the Icelandic banks during the run-up to the crisis. These banks were having a hard time finding funding. My employer implemented structured funding with them, which embedded very profitable derivatives transactions while offloading the credit risk to other institutions—typically hedge funds. Would the bank have looked as hard to find investors if not for the derivative transaction? I doubt it.
Finally, another example of the violation of the interest of the client was the valuation reports for over-the-counter derivative transactions. Over-the-counter derivatives are not listed on an exchange; therefore, it is hard for the client to track the evolution of its value. The valuation report indicates the amount of money that the client would receive or pay to exit the derivative transaction. In general, we tried not to provide the client with these reports, as we would prefer to maintain the control of this information. However, certain clients required it, and we had to comply with their requests. Mark-to-market reports also generated an additional problem; if the client received a mark-to-market report shortly after the transaction, it could back out of the markup that we had put in the transaction. To prevent this, we would try to delay the first valuation report so that the market moved. Another alternative was to spread the markup over a long period—for instance, over 12 months—so that it would appear progressively.
Investment banks have extensive processes in place to limit the risk appetite of front-office employees. However, the same challenge of the asymmetry of information (and asymmetry in the power of incentives) occurs internally. Credit officers, risk officers, and compliance officers typically possess a less quantitative background as the front office people they are facing; otherwise, they would likely be working there. They also rely on the front office to obtain certain information, such as the rationale of the transaction. All of this leads to front-office staff internally developing information asymmetry. I
personally spent hours doing so, “sweet-talking” credit officers into agreeing to our proposed transaction. The classic argument of putting the client first was also largely relied upon, as we would extensively argue that the transaction was in the client’s interest or that it had been requested by the client. I sometimes went further: for instance, on the recommendation of my head of desk, I asserted to credit officers that we were not making any money on a given transaction, although this was incorrect (the markup was close to a million euros). This supported the argument that the transaction was in the client’s interest. My superior had assured me that the credit officers did not have direct access to the markup reports and that they only relied on the information we were providing them.
Compliance was putting a limit on the markup of transactions of derivatives: we could not go over 10 percent of the notional. However, in one occurrence, the client agreed to a transaction that was over this threshold. Instead of improving the pricing, we broke down the transaction into two components in the booking system for the same notional. Each of these transactions, therefore, met the compliance criteria.