A decade ago, the introduction of electronic trading, in conjunction with straight-through processing, increased the efficiency of hedging foreign exchange (FX) exposures. Electronic trading platforms also make it possible to hedge smaller exposures, and they enable companies to hedge exposures in real time as they arise, instead of on a monthly or quarterly basis. As a result, a large portion of FX hedging is conducted on electronic platforms. In fact, the proportion of FX trades executed via electronic platform is currently 46 percent, up from 20 percent in 2004, and the volume of electronic FX trades increased by almost 500 percent over the same period (see Figure 1, below).
Some corporations now use electronic execution to supplement their voice execution, but many use a digital platform for 100 percent of their hedging program. The justifications for using only electronic trading are generally that the platform offers superior pricing and that the company’s hedging strategy and products are “plain vanilla.” In extreme cases, hedging via the electronic platform is company policy and voice execution requires approval as an exception. This can be problematic, as inefficiencies can arise when a company uses only electronic platforms for FX trading.
Transition to Digital FX
During the early to mid 1990s, participants in the FX market traded primarily via phone or phone-based technology. Generally, a client would call either its bank or a broker to get a two-side market (bid and offer simultaneously) prior to executing. By having the bank quote both sides of the market before trade execution, the company achieved price discovery. Post-execution, the bank would hedge its own risk by calling other dealers directly or calling an interbank broker. So banks used telephone lines both to execute trades with clients and to distribute risk via the interbank market.
Through the late 1990s, end users—such as corporations and hedge funds—continued to trade predominantly via voice, but the interbank market began shifting to electronic broker platforms to obtain better price discovery and to distribute risk more efficiently. Previously, interbank trades had been large and less frequent. The electronic interbank market allowed banks to hedge small end-user transactions quickly. It also opened up the market for small dealers by providing a live stream of best bid and best offer, enabling them to hedge risk.
In the early 2000s, multiple electronic platforms were developed for end users. These gave end users the ability to execute via a single-dealer or multi-dealer platform. A single-dealer platform allows clients to receive quotes from, and execute with, one dealer. The multi-dealer platforms connect end users to multiple dealers for quotes and transactions. Unlike the single-dealer platforms, multi-dealer platforms make it possible for end users to run live auctions when executing trades. In a matter of seconds, an end user can obtain multiple bids from a group of dealers they select. After obtaining the bids, the end user chooses which dealer to award the trade, whether or not it offered the best price in the auction.
For both new and experienced hedgers, the increased speed and transparency of electronic platforms generally translated into lower administrative and execution costs. Electronic trading reduces execution time per trade, booking errors, and misquotes. The steady stream of price quotes, coupled with the ability to run auctions, increased transparency. Straight-through processing lowers the administrative burden of inputting trades into internal systems, settling payments, and making accounting entries. The exact savings available through electronic trading differ from user to user, but in general smaller end users that lacked infrastructure reaped the largest savings relative to the size of their exposure.
Gaining Efficiency but Losing Guidance
Something important was lost as companies begin relying more on electronic trading platforms and communicating less with their foreign exchange salespeople. FX salespeople add value on an ongoing basis by explaining best practices and helping end users adapt their hedging strategies to suit the always-shifting market environment. In lieu of salespeople, electronic end users began turning to online research reports, which can often be accessed via the electronic trading platforms. But a research report cannot draw on years of professional experience to recommend the right hedging strategy for each end user’s unique situation.
Moreover, this shift among FX end users has led banks and brokerage firms to place less value on their salespeople. Generally, as FX salespeople have less direct contact with clients, senior management at the bank or brokerage firm begins to perceive them as less integral to revenue. Eventually, management reduces head count or hires less-experienced salespeople. Salespeople with less expertise may not have the long-term FX perspective that is vital to developing a hedging program. And a smaller sales team may be unable to adequately cover all clients, so they may prioritize those that use voice trading or a single-dealer platform, since those platforms more directly link their work with revenue than do electronic trades.
At the same time the number of experienced salespeople in the FX market is declining, the importance of hedging is increasing because foreign earnings are becoming a larger part of net income for U.S.-based businesses. For example, in 2011 foreign sales within the consumer staples industry were around 47 percent, up from 37 percent in 2005. Some banks have tried to supplement deficiencies in their FX sales staff by introducing separate groups to provide strategy or quantitative analysis consulting.
Today, a typical FX salesperson focuses on execution, not strategy. The loss of salespeople who proactively provide strategy advice has left end users reliant on FX programs that they developed years ago. Many continue to use hedging strategies built for market environments in which the euro was stable, interest rates were high, and volatility was lower—an approach that may prove inefficient or even detrimental. For example, many companies continue to roll their hedges on a monthly basis. They developed this strategy as a way to save money on forward points. However, in the current interest rate environment, forward points are small in many of the G10 currencies. Rolling hedges on a monthly basis produces almost no forward-point savings right now, while it increases the amount of bid/offer paid. Companies could reduce their hedging costs by more than 50 percent if they extended the roll period from monthly to bi-monthly.
End users that execute large volumes electronically without proper advice may face even greater losses by moving the market against themselves. Slippage, the difference between an estimated transaction cost and the amount actually paid, can increase quickly without warning when there is a material change in market conditions. A lack of liquidity or an error in an asset management firm’s trading platform can cause sizeable market swings. For example, on May 6, 2010, in the midst of the “Flash Crash,” many companies continued executing FX transactions via electronic platforms. The euro declined by 1.6 percent that day, then immediately rebounded the next day. An FX salesperson can help companies anticipate and prepare for these types of market idiosyncrasies.
Rise of the Algorithm
Not all companies are created equal in the eyes of electronic FX trading platforms. In the equity markets, a known, fixed fee is charged for execution. By contrast, FX trading involves no mandates to provide good execution. The goal of dealers is to take the largest bite possible. Thus, the price offered to a multinational with an AA credit rating is not the same price offered to a BB-rated multinational. And smaller, high-yield companies may achieve better pricing than large multinationals.
To support the large volume of electronic trading, dealers developed algorithms that automatically price trades. Over time, the algorithms have evolved. Today’s algorithms provide differentiated pricing to maximize profit based on deal size, counterparty, tenor, currency pair, time of day, and previous win ratio. For example, the algorithm may detect a pattern in which a counterparty awards business to dealers on the basis of who offers the tightest bid/offer spread. Once it recognizes this pattern, the algorithm will offer tight bid/offer spreads on a poor market price. Dealers have teams dedicated to building smarter, faster, and more discriminating algorithms that continuously improve their understanding of the market and their counterparties. They study the behavior of clients and look for ways to make incremental revenue.
Algorithms also use information gathered from the relationship manager and FX salespeople. Consider Multinational ABC, which trades almost exclusively via an electronic multi-bank platform. Periodically Investment Bank Z’s FX salesperson receives a report detailing volume, win/loss ratios, hedge maturities, and revenue associated with ABC, information the bank gathers directly from ABC, from public financials, and/or from the relationship manager. Based on this data, the salesperson may suggest adjustments in how Bank Z’s FX algorithm responds to Multinational ABC’s transactions. If the analysis reveals that Bank Z is losing more trades than it wins, the salesperson might suggest tighter margins. On the other hand, if the salesperson feels Bank Z is winning too many trades, he or she may recommend expanding the bank’s margins with Multinational ABC.
Apart from Bank Z’s analysis of its win/loss ratio with Multinational ABC, the company’s stated goals may negatively affect what prices the algorithm offers. If ABC proactively distributes its FX hedges among a group of relationship banks, sometimes awarding trades to banks that don’t offer the lowest price and other times removing banks from the bidding process when they exceed pre-set concentration limits, Bank Z’s FX salesperson will advise the managers of the electronic platform to keep the bank’s margins higher. Bank Z will aim for its bids to come in second place because eventually ABC will award Bank Z a share of its business, and at a higher margin than the first-place bank achieves. If every bank uses this same strategy, it will quickly erode any perceived cost efficiencies for ABC of using an electronic bid process.
Voice-Based Trades as Carrot
Voice execution can be a direct substitute for electronic trading, or it can be used as a reward for the dealer that consistently provides the best electronic execution. More than $1 trillion in FX trades occur daily in North America, making this the most liquid of all markets. Live exchange rates are available via a streaming service at a reasonable cost. And foreign exchange is one of the few markets in which it’s standard practice to ask for bid and offer before providing details of the trade. This is how dealers achieved a fair price before electronic platforms existed. With the right systems or independent advice, voice trading can achieve the same level of execution efficiency as an electronic platform.
An indication that voice execution is on par with electronic is that many sophisticated end users continue to use voice platforms as their primary means of execution. The breakdown between voice and electronic FX trades for non-financial companies is about 50:50. If a company executes more than 75 percent (on a notional basis) of its FX trades electronically, then it may want to reconsider its FX hedging strategy and execution protocol. A combination of voice and electronic trades may result in the lowest hedging cost possible, by combining tighter spreads with improved advice.
When they use voice trading as a carrot, end users knowingly pay a wider spread to their dealers as a reward for their electronic execution. This strategy capitalizes on the fact that FX salespeople have meaningful input into the algorithms—but they prefer voice execution because of its direct link to their performance, which is important for their own promotion and compensation. While bank management is indifferent to (or, arguably, even prefers) electronic trading, the average salesperson prefers a dollar made via voice over one earned electronically. For hedging end users, understanding and leveraging the motivators of an individual salesperson is instrumental in minimizing hedging costs in the foreign exchange market.
Joseph C. Lewis is a director with financial risk management consultancy JCRA Financial. He has over 12 years’ experience providing risk management advice on interest rates, foreign exchange, and commodities to corporations, real estate, and private equity funds.