A Discretionary Approach to Currency Investing

07 Apr 2015

Currency investing started in earnest with the introduction of floating exchange rates after the breakdown of the Bretton Woods system in 1973. Although with a history spanning a period of over 40 years, the specialty of currency investing is still relatively new compared to the professional practice of investing in stocks and bonds. In this chapter, we give a brief history of how Fischer Francis Trees & Watts (FFTW) – an institutional fixed income firm – entered the field of currency investing when it began managing global bond portfolios in the late 1980s. In the next two sections, we discuss the early evolution of FFTW’s currency process in the context of the distinct objectives of risk reduction using passive or dynamic hedging, and return enhancement, where the focus is on pure alpha generation. We then examine evidence (following Pojarliev and Levich, 2008) showing that currency managers, including FFTW, have added value as a group over long periods, and that their excess returns can be explained as consisting of true alpha – generated using a primarily discretionary approach in the case of FFTW – and currency beta factors of trend, carry and value, generated typically using systematic models. From this evidence of value-added using both discretionary and systematic approaches, we discuss these two different styles of currency investing.

The chapter then reviews early research; findings that currency models based on fundamental economic variables have not been particularly successful. This failure of fundamental approaches has led to the use of models based on statistical price measures and other non-economic, technical, inputs. We describe the general challenge in building systematic models, and the difficulty of living with such models in practice. This is followed by a look at discretionary processes in currency and the features of currency markets that we believe make it possible to anticipate capital flows, and thereby predict currency movements using judgment. We finish our comparison of the systematic and discretionary approaches to currency investing by discussing some of the pros and cons of each style.

The next section turns to FFTW’s overall approach to currency investing, a blend of a judgment-only discretionary approach and the use of systematic models to capture beta factors that include trend and carry. We emphasise the importance of risk control, both with discretionary and systematic styles, due to the ever-shifting nature of currency dynamics. This is followed by an analysis of the returns and correlations of FFTW’s primarily discretionary process, naïve currency beta factor styles and major asset market indices, including stocks, bonds, hedge funds and commodities. We find that FFTW’s judgment-only returns, and the trend and value currency beta factors, have very low correlations with major indices. In the case of FFTW’s approach, we also find that currency returns are positive in 2008, the year of the Lehman crisis, as well as in 2009, the year of the economic and stock market rebound. Our correlation and return analyses – be it of the discretionary process exemplified by FFTW, or systematic approaches represented by naïve beta factors – strongly suggest that investors should consider allocating to currency investing as a separate asset, alongside stocks, bonds and other major asset classes.



FFTW was launched in 1972 by Richard Fischer, Stephen Francis, James Trees and John Watts, who had worked together at Brown Brothers Harriman, and had the pioneering idea of applying modern portfolio theory to the active management of fixed income portfolios. A substantial industry of active portfolio management for equities had already been in place for some time, and the new discipline of modern finance (which developed after the seminal work of Markowitz, 1952) had led to a new understanding of risk and return in the world of equities, with practical implications for the industry of active equity investing. The newly founded form was a partnership that focused solely on fixed income and actively managed institutional bond portfolios for total return – a novel concept in 1972, when bonds were mostly held to maturity in the trust departments of major banks.

After growing for more than a dozen years as a US dollar-based institutional fixed income manager, FFTW decided to expand its active management capabilities to global bonds from late 1989. The decision to go into global fixed income was a natural extension of two characteristics of FFTW’s business since its founding in 1972. The first important factor was that, from the start, FFTW had two complementary businesses in fixed income: (i) an active fixed income investment management business for institutional clients; and (ii) a proprietary trading activity that was professionally managed by a separate investment team to earn an attractive return on the partnership’s capital. The existence of proprietary trading allowed FFTW to experiment using its own capital with instruments and approaches that were not commonly used in domestic fixed income portfolios at the time. Among these novel approaches and instruments were leverage, shorting, derivatives, options and foreign exchange.

The second important factor was that, although FFTW only managed US dollar fixed income in its institutional business line for the first 15 years, many of its clients – even as early as the mid-1970s – were institutions such as central banks, international of financial organisations and sovereign wealth funds who had given FFTW US dollar mandates for part of their US dollar reserve or pension portfolios. After a decade of relationship with FFTW, a number of these of financial institutional clients were interested to have FFTW expand its active fixed income activity into global fixed income to cover their non-US bonds.

It is in this context that FFTW seized the opportunity to develop an expertise in global fixed income by opening a London office in 1989, and started to manage active global sovereign bond mandates for a European official institution and two European pension plans. The benchmarks for these active mandates included the recently launched Salomon World Government Bond index (now WGBI), which represented a multicurrency benchmark. With the launch of the global bond product at FFTW, a new investment process with new concepts and tools was also in order. Whereas domestic US fixed income was familiar territory for FFTW and competitor fixed income managers, global fixed income was a new field requiring a more formal approach to the identification and budgeting of various active decision risks. In domestic fixed income, the notion of separating duration, yield curve and sector decisions into individual investment activities, each with their own processes, expected returns and volatilities, was not the norm. However, with global fixed income it soon became clear that, in addition to the various components of bond risk inherent in domestic fixed income, there was a distinct new dimension of decision risk in the active management of currencies.


Download to read more