This article argues that price-based strategies, such as trend following, could prove more reliable than systematic strategies based on fundamentals, given the uncertain macroeconomic outlook. It then goes on to highlight less widely-discussed properties of trend following that may be favourable in the current environment.
Lockdowns introduced to stop the spread of Covid-19 brought most of the global economy to a standstill in March 2020, with simultaneous supply and demand shocks leading to large moves in unemployment and other economic indicators (see Figure 1). This prompted sweeping rounds of monetary and fiscal intervention from central banks and governments around the world. The last time investors faced a similar situation was the 1940s, when large parts of the global economy were shut or nationalised as a result of World War II.
Market reactions to the deteriorating economic outlook were dramatic. The MSCI World Index slid by a third between February 12 and March 23, US 10-year Treasury note yields fell below 1% for the first time on March 3, and front-month WTI crude futures briefly turned negative on April 20.
Source: Federal Reserve Bank of St Louis (A and C), Bloomberg (B), as at 5 May 2020.
The sudden and material changes in the global economy have made forecasting asset prices difficult. Take equities, for example. Future corporate profits will be highly sensitive to the easing of restrictions and the terms of ongoing government intervention. As many as 79 S&P 500 companies had withdrawn profit guidance for 2020 altogether during first-quarter earnings season (by April 30).
The uncertainty in the economic outlook is visible in the range of 2020 forecasts published by bank strategists and economists during April, which are far more dispersed than the full-year forecasts made at the same point in previous years (Figure 2: A and B).
There is also evidence of consensus forecasts predicting very different outcomes to those implied by market prices. By way of example, analysts predict a swift rebound in dividend pay-outs from the constituent companies of the Euro Stoxx 50 index after 2020, while futures markets are pricing in a deeper and more sustained drop (Figure 2: C).
Source: Winton, Bloomberg and Wall Street Journal, as at 30 April 2020. (A) Bloomberg Bank Strategist Survey (latest survey available on April 30 each year); (B) Wall Street Journal Economic Forecasting Survey (April 1 survey each year); (C) Bloomberg consensus analyst forecasts and EUREX Euro Stoxx 50 Index Dividend Futures.
This backdrop poses a challenge for all investment managers, but particularly those that seek to predict asset prices systematically using models based on fundamental relationships. Data sourced from financial statements and government statistics bureaux are inherently backward-looking, which means the information they contain may not be relevant for navigating the period ahead. And while consensus forecasts factor in the impact of the coronavirus, they are based on disparate assumptions about the responses of governments and populations, and the effectiveness of those responses. Investment managers must exercise extreme care when assessing whether models that use such data remain valid.
Strategies that rely only on intrinsic market data – such as prices and trading volumes – avoid the uncertainty associated with fundamental data. This is because prices are reliable metrics as long as markets remain liquid and assets can be marked to market. Models based on fundamental data, meanwhile, attempt to quantify an unknown property of the world or an asset, in order to forecast asset prices.
Trend following, which Winton has been trading since 1997, is the quintessential price-based investment strategy. Put simply, the strategy buys assets that have increased in value, on the assumption that they will continue to increase in value, and sells assets that have fallen in value, on the basis that their prices will continue to fall.
This approach to investment management has proven to be highly profitable since managers began explicitly adopting it in the 1970s and 1980s. Trend following has delivered attractive risk-adjusted returns that have been lowly correlated with other assets and investment strategies over the long term.
The strategy’s systematic and highly diversified approach − which is applied to all major liquid financial assets, currencies and commodities − is essential for ensuring that it works. At any given point in time, the strategy will be long or short each asset in its trading universe, regardless of analyst forecasts, market narratives and economic or financial theory. Trend following does not seek to predict where and when trends will emerge, but rather to profit from them wherever and whenever they do.
In the remainder of this article, we look at how trend following is able to perform well when market forecasts are unreliable; its tendency to take contrarian positions; how significant shifts in market consensus can create opportunities for the strategy; and whether now is an appropriate time to make an allocation, given the strategy’s recent strong performance.
Trend following has a track record of performing well when fundamental forecasts have been poor. Figure 3 explores four examples of this, by plotting forecasts made by economists and bank strategists against the relevant markets. Below each chart, we show a trend-following signal and its simulated P&L.
Source: Winton, Wall Street Journal Economic Forecasting Survey, Bloomberg Bank Strategist Survey, as at 30 April 2020. The starting point of each line indicates the yield/close price on the date the forecast was published. The P&Ls are simulated and do not represent the results of actual trading. The accompanying disclaimers at the bottom of this page should be read carefully when reviewing this data. No representation is being made that any account will or is likely to achieve profits or losses similar to those being shown.
The examples in Figure 3 show how trend following’s systematic approach to investing can provide diversification from strategies that are reliant on market forecasts. They also demonstrate how the strategy can, counterintuitively, behave in a contrarian manner, by taking positions that conflict with market narratives and fundamentals.
Trend following is often perceived as a strategy that simply “follows the herd”. However, this metaphor is inadequate, as the strategy has been shown to be particularly profitable precisely when it holds positions that – at least, initially – are at odds with the prevailing market consensus.
Price trends tend to emerge in markets as the consensus position gradually shifts. And trend-following algorithms − dispassionately adjusting their positions each day in response to new information – do not face the dilemmas of discretionary traders over when to throw in the towel or change course.
The global lockdown has disrupted trade patterns, cross-border supply chains and international travel. It could go on to transform many other areas of economic activity, from the nature of work and the provision of healthcare, to consumer behaviour and the role of government in the economy. Predicting winners and losers across markets and timeframes is difficult ex ante, but the beauty of trend following is that one need not even try.
The reason is that the strategy is applied systematically to a broad range of markets, industries and asset classes, with the direction and size of positions determined by the strength of the price trend and its volatility. As a result, the larger the price movements relative to its volatility – a measure of an asset’s “trendiness” – the larger the position the strategy will take in that asset, and the larger the returns are likely to be.
These strategy characteristics are evident in the data. In Figure 4, we simulate the performance of a trend-following system on more than 100 futures markets over 40 years, or as far as the data allows. We then plot these returns against the risk-adjusted size of annual market returns, positive or negative, for each year. And as the grey markers show, the strategy makes large numbers of individual bets each year, which tend to do better in markets where there are large risk-adjusted moves, relative to volatility.
To help put this performance in context, the averages during 2019 for five sector groupings are highlighted in purple – and, for comparison, the same results for 2008 are in blue, a particularly good year for trend following.
Source: Winton, as at 31 December. These results are simulated and do not represent actual trading. The accompanying disclaimers at the bottom of this page should be read carefully when reviewing this data. No representation is being made that any account will or is likely to achieve profits or losses similar to those being shown.
There were big market moves across all five sector groupings in 2008, hence the strong trend-following results that year. By contrast, the strategy’s performance in 2019 was driven solely by an uptrend in fixed income. Trend following failed to profit from large risk-adjusted moves in equity indices, while currency and commodity markets were directionless overall.
After largely trading sideways since 2015, trend followers generally performed well in 2019, with the SG Trend Index rising 9.2%. The strategy has continued to deliver returns in the first quarter of 2020, with the same index rising 2.3%, as the MSCI World Index fell 21.1%. While performance-chasing is a widely observed phenomenon in investment management, some may question whether it is wise to invest in trend following after its recent strong run.
The analysis in Figure 5 shows how recent performance says little about how the strategy will perform in subsequent periods. Each blue marker plots the return of a simulated trend-following strategy over a period against its return in the previous period. If periods of negative performance tended to follow positive performance (and vice versa), the markers would be oriented from the bottom right of the plots to the top left.
Source: Winton, Bloomberg, as at 28 April 2020. These results are simulated and do not represent actual trading. The accompanying disclaimers at the bottom of this page should be read carefully when reviewing this data. No representation is being made that any account will or is likely to achieve profits or losses similar to those being shown.
The lack of such a pattern suggests that attempting to time trend following with a tactical position may be difficult. Investors may therefore be better off holding the strategy as a long-term strategic allocation.
The benefits of systematic trend following − such as its ability to make money in rising and falling markets or provide a source of long-term diversification for equities and bonds − are now well understood by the institutional investment community. In this article, we have highlighted less widely-discussed properties of the strategy that may prove favourable in the current environment.
For instance, the fact that trend following does not use market fundamentals, or forecasts based upon such data, should reassure current and prospective investors. These data points are unlikely to be reliable strategy inputs, given the upheaval and uncertain outlook resulting from lockdowns in the world’s largest economies.
And, by ignoring prevailing market narratives and fundamentals, trend-following strategies will be able to maintain contrarian positions that could provide valuable diversification from other strategy allocations in investors’ portfolios.
These benefits do not just relate to the immediate future. As the situation eventually stabilises, the global pandemic will inevitably result in winners and losers across industries and countries. It may be impossible to predict with any certainty the extent to which individual markets will exhibit more trends in the coming years than in recent times. But if the pandemic-inspired economic realignment produces such an environment, trend followers will be well placed to profit.
The information herein does not constitute an offer to sell or the solicitation of an offer to buy any securities.
This document contains forward-looking statements that are based on current indicators, expectations and assumptions that Winton believes to be reasonable but are subject to a wide range of risks and uncertainties and, therefore, there can be no assurance that actual results will not differ from those expressed or implied by such forward-looking statements. These statements speak only as of the date on which they are made.
This document contains simulated or hypothetical performance results that have certain inherent limitations. Unlike the results shown in an actual performance record, these results do not represent actual trading. Also, because these trades have not actually been executed, these results may have under- or over-compensated for the impact, if any, of certain market factors, such as lack of liquidity and cannot completely account for the impact of financial risk in actual trading. There are numerous other factors related to the markets in general or to the implementation of any specific trading program which cannot be fully accounted for in the preparation of hypothetical performance results and all of which can adversely affect actual trading results. Simulated or hypothetical trading programs in general are also subject to the fact that they are designed with the benefit of hindsight. No representation is being made that any investment will or is likely to achieve profits or losses similar to those being shown.