اولین سرمایه گذار جهانی بازارهای نوظهور : تراست سرمایه گذاری خارجی و مستعمراتی 1880-1913

The Foreign and Colonial Investment Trust is the oldest surviving closed end fund, having been established in 1868. Its early success and emulation were related to its identification of a missing market – the provision of a wholesale diversified vehicle for the investing public. This paper is a micro-study of this leading investment trust during the First Era of financial globalisation. The history of this flagship fund over more than three decades provides an insight into the relative success of this financial innovation as well as into the risk and returns of investing in emerging markets over a century ago.

The Foreign and Colonial Investment Trust (FCIT) is the oldest surviving closed end fund in the world today. Improving upon an idea first adopted by the Dutch in the 18th century (Rouwenhorst, 2005), the trust started in 1868 fully half a century before such funds first appeared in the US. Established as the Foreign and Colonial Government Trust, it was substantially reorganised a decade later. Our analysis of the annual portfolios from 1879, when its shares first became listed on the London Stock Exchange (LSE), until 1913 provides an insight into how one sophisticated investor approached the rapidly expanding world of international investment during the First Era of Globalisation (O'Rourke and Williamson, 1999 and Obstfeld and Taylor, 2004). This paper is therefore a micro-study of one of the pioneering investment institutions that opened up the way to investment in emerging markets in the three decades up to World War I. Compared with the very extensive literature on technological innovation, financial innovation is still a field with many unanswered questions. Most innovations in the financial sector are similar to general purpose technologies that generate benefits for the immediate adopters, but can also have widespread impacts on the whole economy. Some of the latter take the form of externalities, often negative, which in turn invites discussion as to the net social benefits of the financial sector (Levine, 2005 and Litan, 2010). Among the criticisms levelled at financial innovations is their potential to create ‘systemic risk’, harm consumers and waste resources due to rent-seeking behaviour by the innovators themselves. Recent reviews of the finance literature therefore make a plea for more study of the costs and benefits of financial innovation, its diffusion process and, especially, its origins (Frame and White, 2004 and Lerner and Tufano, 2011). Financial history is an obvious means to address this gap. In this spirit, our paper provides the first quantitative account of the origins of the investment trust industry. And, based on primary sources, seeks to answer several of the questions about the origin, diffusion and benefits of this mutualised form of investment. The contribution of this study is fourfold. First, the FCIT successfully identified a missing market – that for wholesale investment in diversified portfolios by the general public – at a time when domestic securities were yielding historically low returns. Over the period 1880 to 1913, FCIT delivered attractive risk-adjusted returns to the investing public at very modest cost. The fund's investments averaged a total return of 5.3% per annum, well in excess of the 2.2% return on British Consols with a better risk-return trade-off. Furthermore, in only two years did FCIT record a negative return and the maximum drawdown was only 8% in 1890. The investment trust model gained in popularity over the period, such that by 1913 there were 61 investment trusts quoted on the Stock Exchange, with a combined market capitalization of close to £90 million. In documenting the development of the investment trust sector, the important distinction needs to be drawn between the original “average trusts,” pioneered by FCIT and focused on delivering long returns with low leverage, and the speculative fringe of “financial trusts” which emerged in the 1880s and quickly failed during the Baring crisis. US investors in investment trusts went through a similar costly experience in the late 1920s and early 1930s (Delong and Shleifer, 1992). This experience resulted in their marginalization and the rise of the open end mutual fund (Fink, 2008). Britain ultimately followed where the US led over the rest of the twentieth century. In this paper, we argue that the history of mutual funds post-1929 has overshadowed some of the more desirable features of the original closed end fund innovation as represented by FCIT. Second, FCIT maintained its investment focus on emerging markets throughout the period. From its beginnings as a portfolio of “well-selected Government Stocks” it added colonial government securities and then US railroad stocks. After 1890 the fund moved to a 90% exposure to the New World outside Europe. A similar model of institutional investment had been pioneered in the Netherlands in the late eighteenth century (Rouwenhorst, 2005). However, whilst Dutch mutual funds substantially concentrated their investments in the US during the eighteenth century, FCIT took full advantage of the broader range of investment opportunities available by the late nineteenth century to construct a globally diversified portfolio. Hence, a study of FCIT provides new insights into the risks and opportunities confronting international investors over a period of 33 years by a rapidly expanding developing world and invites comparison with recent emerging market investment since its re-emergence in the early 1990s. Third, we provide early evidence on the existence of a closed end fund discount which has in the later twentieth century been considered a “puzzle” (Lee et al., 1991) given that a well-capitalized investor might be expected to exploit the law of one price, acquire such a fund and liquidate its investments. To the best of our knowledge, the only prior historical study of the closed end fund discount was undertaken by DeLong and Shleifer (1992). The authors found that the highly-leveraged US funds traded at a premium up to 1929, before suffering heavy losses and trading at substantial discounts in the 1930s. By tracing how FCIT's deferred shares traded relative to the net asset value (NAV) of the underlying investments, we estimate a level of discount for the FCIT which is not out of line with investor experience later in the twentieth a century. One explanation for this discount is, similar to today, the exposure to illiquid securities. In addition, fluctuations in the discount around the time of the 1890 Baring crisis appear to reflect the ebb and flow of investor sentiment, also in keeping with experience later in the 20th century. Last, from its inception FCIT pursued a patient buy-and-hold investment approach rather than a high frequency trading approach seeking to exploit short-term mispricing of securities. This is most apparent in the very low turnover of its portfolio compared to that of modern institutional investors. That FCIT was able to adopt this approach is due at least in part to its choice of the closed-end fund structure. Unlike open-end mutual funds which emerged on both sides of the Atlantic in the 1930s and offered shareholders the facility to redeem units on a daily basis, this structure diverted the pressure of shareholders withdrawing their funds into the secondary trading of its shares on the LSE and left the managers free to hold onto their investment positions, particularly during financial crises. Our analysis of FCIT's security trading during the two most significant crises of this era, 1890–93 and 1907, suggests that the managers displayed no evidence of contributing to any emerging market contagion. Such evidence is pertinent to the evaluation of the social welfare benefits of the investment trust industry, which may have differed from other investment institutions, often perceived as a source of ‘systemic risk’ and ‘contagion’ in financial markets. This paper sets out to draw more general conclusions from the study of a single investor in the manner of Temin and Voth (2004) and Chambers and Dimson (2013). We argue that FCIT was representative of those features of the “average” investment trust which made it a successful innovation, namely, the provision with daily liquidity of risk-adjusted returns superior to those on domestic assets via an internationally diversified portfolio assembled at low cost via a patient buy-and-hold investment approach. The speculative financial trusts which populated the sector alongside average trusts in 1880s Britain, and in the US of the 1920s, did not have these features and were relatively short-lived. Even though FCIT's remarkable longevity should raise concerns about survivorship bias, the evidence shows that the fund was ‘average’ not only in its investment strategy, but also in its performance. FCIT does not seem to have benefited from any first-mover advantage or fund manager skill compared to its peers. The rest of the paper begins with a description of the origins of the investment trust industry, and FCIT in particular, in the context of the rise of international capital markets prior to 1914. Section 3 then describes our data and presents summary statistics on the FCIT portfolios. Section 4 discusses the extent of diversification achieved by the FCIT portfolios. In Section 5 we first review the performance based on both NAV and share price and estimate the discount to NAV. We also benchmark the performance of FCIT and discuss contagion in the context of how the fund invested during the two major financial crises of 1890 and 1907. Section 6 concludes.

There could be no higher testament to FCIT's attractions than the fact that this was the only investment trust share which John Maynard Keynes included in the security portfolio he managed for his father before WW1. FCIT stands as the forerunner of a highly successful financial innovation, based on the idea of providing the average investor with the opportunity to invest in a globally diversified portfolio at an extremely low cost. Back in the late 19th century, when it would be too costly both in time and money for individual investors to try and replicate such a portfolio, this proved to be a highly attractive and convenient solution for the majority of the investment public. The investment trust structure permitted the FCIT to exploit the benefits of a long-term investment horizon and to pursue its buy-and-hold investment approach. This in turn allowed the board to make heavy investments in emerging market bonds and in the American continent in particular, an investment approach which paid off handsomely. The fund's NAV averaged returns in excess of British Consols and of the global ex-UK benchmark whilst also offering a better risk-return trade-off. The trust's deferred shares exceeded its NAV performance thanks to the leverage provided by the issue of preference shares. However, as our peer fund performance comparison showed, FCIT's performance whilst creditable was not particularly outstanding and there were other average trusts that also performed well. In one respect FCIT did appear to be outstanding and that was disclosure. Consequently, we were able to estimate the NAV of the underlying investments and showed that the level of discount to the NAV at which the deferred shares traded was not out of line with what investors a century later experienced and most probably reflected the exposure to illiquid securities in the portfolio. A straightforward comparison with today is offered by the Emerging Markets Composite Index (EMBI), launched in 1989 by J. P. Morgan, and now generally recognised as the emerging market debt benchmark. From inception in 1993 through the end of 2010, the EMBI returned 11.7% p.a. in US dollars compared to 8.6% for US Treasury 30-year Bonds was slightly lower at 15.8% versus 16.8% for the US long bond. Hence, in the modern era spanning a period of 17 years, emerging market debt generated a US dollar return of 3.1% p.a. in excess of US government bonds, a premium which is identical to that returned by FCIT over Consols in sterling pounds in the pre-WW1 era spanning a period almost twice as long. Moreover, the FCIT return premium was earned with much lower volatility, only 4.7% compared to EMBI volatility of 15.8% and US Treasury bond volatility of 16.8% in the recent past. As a result, the Sharpe Ratio was higher at 0.67 in the earlier period compared to 0.53 today. The success of the FCIT model was imitated first in Britain and then, after the war, in the US. This model created both followers and epigones. The latter exploited the closed end fund structure to pursue excessive leverage, speculative investments and even pyramiding. The greater risk taken on by these “financial” trusts inevitably resulted in more instability and greater failure rates than the original and more conservatively managed “average” trusts. Among the followers of the FCIT example were the other 37 trusts that not only survived the Barings crisis but remained in operation continuously until the War. In the US, closed end funds were subject to considerable criticism in the wake of the 1929 stock market crash and were superseded by the new open end mutual funds. The latter type of fund also came to dominate investment trusts in Britain through the remainder of the last century. This study on the origins and development of the investment trust cautions against overarching negative narratives of the closed end fund innovation drawing on the speculative excesses of the US during the 1920s or Britain in the 1880s.. Whilst the closed end fund structure was suffered from fluctuations in its discount to NAV, an added source of risk for the fund investor, at the same time it permitted average trust managers such as FCIT to undertake the patient buy-and-hold investment approach free from fund flow pressures and was therefore particularly suitable to a less liquid asset class such as emerging markets. Mauro et al. (2006) suggest that the institutionalisation of investment activity over the 20th century is crucial in explaining the greater incidence of emerging markets contagion in the Second Era compared to the First. However, the trading behaviour of FCIT during the two major financial crises of 1890 and 1907 qualifies this generalisation. The evidence suggests that the introduction of the investment trust did not necessarily come at the cost of greater contagion in emerging markets. The implication of this study is that not all investment institutions are created alike.