Performance and the Short and Long-term
The European Commission is expected to produce a Green Paper on long-term investment in December. Long-term investment has been the subject of much academic and official work in recent years; second only to crisis-cure in volume. It may come as a surprise that the topic is rather weakly defined, if at all. The World Economic Forum and OECD have gravitated to the idea of “patient, engaged, and productive capital” but that has the disadvantage that these concepts are themselves difficult and perhaps impossible to measure.
In this note, in addition to drawing out some of the differences between short and long-term investment, we shall propose a ‘straw man’ which is open to measurement, and relate that to the patient and productive concept. Before moving to that, however, we need to consider the reasons why a sequence of short-term optimal outcomes will tend to produce a combined result which is inferior to the optimal long-term result. Among the numerous published studies and reports, this is taken as an article of faith; long-term investment is simply virtuous and short-term reprehensible. This attitude extends across much of the discussion around sustainability, which is ineluctably concerned with the long-term.
The most obvious justification for the pursuit of the long-term is that active management in the short-term incurs recurrent management and trading expenses. This may well hold true in a narrow sense where financial markets are considered simply as the allocation conduits for capital flows from savers to entrepreneurs. The problem is that this is not the sole function of financial markets; they also have roles in capital reallocation, and observed prices some value as signals. Moreover, financial markets may operate as mechanisms facilitating maturity transformation, as complements and competition to the banking system. If there is a problem with the costs of recurrent short-term activity, it most probably lies in the extent to which this activity rewards rent-seeking behaviour within the financial sector.
The economic arguments in favour of long-term investment emphasise the fact that much investment will not be undertaken if a short-term view is taken, and much will be abandoned, with considerable loss of economic resources, before projects reach fruition. The short-term sequential, maturity transformation approach is inherently unstable; even relatively small shocks may topple it and result in costly liquidation.
Of course, many regulations, and particularly the risk-based solvency forms, encourage, or even require short-term approaches. This is not new. Keynes was concerned by the preoccupation with liquidity for investment institutions long ago. In fact, when securities are negotiable in markets, their prices contain an option on liquidity, which is costly; direct investment in non-negotiable securities should produce higher income reflecting this.
There is a missing analysis here, which clearly should be the subject of future research. We should be concerned not just by the volume of savings and investment, but also by the term structure of savings purpose and the term structure of investment demand, and the extent to which financial intermediation mitigates mismatches between these.
The more immediate and practical problem faced by investment managers concerns how the long and the short-term differ from one another, and the consequences of these differences for investment management processes and techniques. Of course, this has been known for rather a long time. The work of Beebower, Brinson and Hood showed that most of a portfolio’s return volatility could be explained by asset allocation – subsequent empirical studies vary but most attribute 80% or more to asset allocation. At a daily or intra-day level almost all performance will be causally due to changes in price. All that this is really telling us is that price performance dominates returns in the short-term. Most performance attribution analysis is to this day conducted in a Beebower framework.
The long-term, however, is a rather different story. The classic study of long-term performance is Dimson, Marsh and Staunton’s work, published as “Triumph of the Optimists”, but also reproduced by many others. In this study of returns since 1900, the real return of 5.5% p.a. achieved by UK equities consists of 4.8% dividend yield, 0.6% growth in dividend yield and just 0.1% change in price. This is as intuitively obvious as the earlier observation that the short-term is dominated by price performance. It also applies to bonds – a bond held to maturity will return the face coupon and principal, and variation will only arise from differences in the achieved reinvestment rates of coupons, which are small effects.
It is clear that investment managers should be using fundamentally different techniques and resource dedication when managing for the short-term than when managing for the long-term. The short-term is predominantly a game against others in markets; it is endogenous. It was well-described by Keynes by analogy to a newspaper beauty parade competition, where it was necessary to select the most attractive as judged by other participants. In recent times, studies show that more than 85% of European equity market price changes have been unrelated to news. A further indication of this short-termism is the fact that 5% or less of market transactions result in a change of registered owner. The long-term, by contrast, is driven by the fundamentals of the company, its earnings and dividends. Dedicating 90% of your analytic resources to price performance may be justified when that accounts for 90% or more of the total, but clearly it isn’t when it accounts for less than 2%.
It is interesting to note that market price volatility is an order of magnitude greater than earnings or dividend volatility. This accounts for one of the issues which has plagued the current generation of performance attribution models; that volatility and value-added do not scale over time in any simple manner. It also lies at the heart of an issue in pension scheme management, where actuarial models make assumptions which are founded in the short-term about the long-term. Simply put, as we move from the short to the long-term, so the total achieved income yield of the portfolio grows in importance relative to the price performance and the result is that the lower income volatility grows in importance. Volatility should be expected to diminish as holding term increases for this reason. Empirical studies confirm this. It is notable that these problems of volatility and value-added scaling were first noticed in bond portfolios, where volatility will diminish as the holding period extends and the term to maturity shortens. It should be emphasised that this is not an argument for mean-reversion in equity market prices.
It also leads us to the straw man, which we wish to propose, for distinguishing the short from the long-term. The income from a security is not sourced in the market, but is specific to the obligor of that security, bond or equity. We shall define the long-term in terms of dependence upon this security-specific source of income and the short-term as dependence upon the market for liquidity, or income through sale. This admits a distinction between investment and speculation. Investment is dependence upon the performance of the contract. In this light, a term savings deposit may be investment. Bonds bought and held to maturity would also be investment, as also would be an equity held solely for the dividend income until dissolution and liquidation of the company.
By contrast, speculation relies upon realisation in a market for its performance. There are actually financial markets which are, by design, entirely speculative in nature. Futures markets are a prime example. Even hedging can be seen as a form of speculation. The objective when hedging is to produce future price returns which offset the price variability of an asset or liability held.
While speculation usually carries pejorative overtones, none is intended here. Indeed, hedging and the earlier-discussed maturity transformation which it admits, clearly can have socially and economically useful benefits.
This distinction between sources of liquidity, income versus market, allows us to distinguish between the short and the long-term. Moreover, we may relate this quantitatively to the holding period. The longer we hold an equity security, then typically the greater the importance of the dividend income received in the total return. This definition of the short and the long-term admits a quantitative scale of the term, and one which is informative about the riskiness of the holding.
It can also be related to the earlier “patient, engaged and productive capital” definition. Patience here is reliance upon contractual terms, and these incomes are precisely related to specific productive output.
Engagement is currently a popular concern, usually rooted in the principal-agent problem between investors and managers. This is often expressed as the difference between ‘exit’ and ‘voice’. For many investors, notably short-term players, it may be more efficient to ‘exit’, by sale in a market, rather than to engage, since engagement is costly. For the long-term investor here, the effect of management rents is directly visible and engagement to rectify these issues more probable as the trade-off can often then be measured. Though quite how one would measure and evaluate management ‘shirking’ is difficult to see. One of the issues for the short-term investor is that this engagement affects prices from which all benefit, though most have not incurred engagement expense; this may place them at a competence disadvantage to other short-term players. However, the long-term investor is unconcerned with market price performance; engagement is matter of self-interest. Engagement, per se, though is a symptom rather than a defining characteristic of the long-term.
This question of the differences between the short-term endogenous game against others and the long-term exogenous game against nature is itself important in terms of investment management incentives. In the former, players are concerned with their position relative to others and price performance matters only to the extent to which it attracts further savers to the market. There is no incentive for such short-term players to want prices to reflect fundamental value, and often many to prefer distortions of this relation. By contrast, the long-term investor wants to see improvements to the fundamental productive output, and is unconcerned by market price behaviour.
Perhaps, the greatest paradox of the short and the long-term lies in the incentives for research and innovation, which is intrinsically a speculative activity. The short-term player will discourage these in pursuit of the immediate, while the long-term investor will encourage them. Ultimately, of course, the benefits of innovations will accrue to consumers, but in the short-term they are a source of supra-normal profits to investors. The far more important point though is that it is research and innovation which are responsible for growth in productivity and productive output. And we have seen precious little of that in recent years.
 Gary P. Brinson, L. Randolph Hood, and Gilbert L. Beebower, Determinants of Portfolio Performance, The Financial Analysts Journal, July/August 1986
 Elroy Dimson, Paul Marsh, and Mike Staunton, Triumph of the Optimists: 101 years of Global Investment Returns, Princeton University Press, 2002. There are annual revisions to these datasets published in conjunction with various investment bank sponsors.