Short-termism is a pandemic that continues to flourish unchecked within financial markets, resulting in significant long-term economic costs to investors. Mark Arnold and Jason Orthman examine the economic costs of excessive short-termism by active fund managers.
Most active equity fund managers that survive over the long-term have investment philosophies focused on exploiting mispricing events, where stock price and long-term intrinsic value have moved apart.
It is the process of price moving towards intrinsic value that provides active managers with alpha. For reasonable quality businesses, intrinsic value should be fairly stable over the short to medium term, and actually trend higher over the very long-term.
Therefore, selling of stocks should normally occur when the price/value gap has closed or another more attractive opportunity is available. The timing of the alpha production is therefore uncertain over short periods of time.
Despite this, most active fund managers are pushed towards short-term thinking.
The majority of funds are in the form of open-ended mandates where investors are able to withdraw some or all of their funds at any time.
Because short-term relative performance can vary significantly above or below the benchmark, and investors are able to withdraw funds at any time, many fund managers undertake activities to reduce short-term relative underperformance risk.
Portfolio managers argue that “failure to achieve acceptable benchmark performance in the short run could lead to large fund withdrawals and their possible dismissal”.
However, periods of short-term underperformance relative to peers or the benchmark are to be expected, even for portfolios that perform well over the long-term.
Funds that produce alpha over a 10-year period can be expected to underperform significantly in quarterly, yearly and even rolling three-year periods.
Increased levels of fund specialisation have also tended to result in more frequent fund performance reviews by asset consultants and trustees, and an automatic increase in the focus on short-term performance.
In theory, it takes between 25 to 40 years for a manager’s track record to reach statistical significance. While this is too long in practice, it highlights the need to focus on longer rolling performance records.
Unfortunately, behavioural biases often result in managers being hired after a short period of strong performance and eliminated after a short period of underperformance. Krehmeyer cites a number of studies where asset managers are fired just before performance improves and hired immediately before performance declines.
A more recent 2008 study by Goyal and Wahal which examined the selection and termination of investment management firms by 3,400 plan sponsors between 1994 and 2003 found that plan sponsors hired investment managers after superior performance but on average, post-hiring excess returns were zero.
Post-firing excess returns were frequently positive and sometimes statistically significant.
One way active fund managers attempt to reduce short-term underperformance risk is by being more active – that is, by undertaking more short-term trading activities. In 1955, the average US fund held its portfolio for seven years; 50 years later, the average stock was held by the average fund for just 11 months.
The average holding period for Australian equities has declined from more than six years in 1986 to under one year currently. An increase in market velocity should correspond to an increase in competitive intensity over shorter time horizons.
Thus, the probability of outperforming is higher by basing investment actions on longer time horizons (five to 10 years rather than six to 12 months).
Miller refers to this as time arbitrage and Gray explains how this can persist, as very few managers are willing to accept the high level of career and business risk required.
The reality is that short-term stock returns are extremely difficult to forecast accurately because of inherent market volatility. Gerlach (2005) found that market volatility cannot be fully explained by fundamental factors.
Schiller (1981) found that market returns are more volatile (uncertain) when measured over short periods of time than when measured over more extended windows.
Looking at Australian equity index returns over rolling 12-month periods from 1969 to 2011 shows that although the average 12-month change in the market PE was close to zero, the spread of return contributions due to changes in historical PE ratios was very wide, and significantly wider than the spread of return contributions for short-term EPS growth (note, the average contribution to short-term returns from earnings growth was also close to zero).
Over longer periods, the absolute range of the PE ratio change stayed fairly similar to its contribution to returns over the short-term.
However, the long-term return contribution from EPS growth showed a significant increase in the contribution to positive market returns.
Both the upper limit of the EPS return range and mean moved significantly higher while the downside contribution was similar to that seen in the 12-month return analysis.
The ability of EPS growth to positively bias long-term stock returns is a key reason why the risk of investing is reduced by taking a longer term approach.
At an aggregate stock market level, buying when PE ratios are high (low) relative to sustainable or normalised earnings will normally result in poor (good) subsequent long-term returns to the investor.
However, over short-term periods of time, the market’s PE ratio is not an accurate predictor of future returns – nor is earnings growth a good predictor of short-term market returns.
These two fundamental factors – PE ratio change and earnings growth – have strong explanatory power for long-term market returns. Long-term earnings growth has a consistently positive influence on future market returns at different initial PE ratio levels, whereas PE ratio change has a positive influence at low initial PE ratio levels and a negative influence at high initial PE ratio levels.
At low initial PE ratio levels, both earnings growth and PE ratio change have a positive influence on future long-term returns. At high initial PE ratio levels, earnings growth continues to have a positive influence on future long-term returns, but PE ratio change has a large negative influence on future returns.
A key benefit of taking a long-term fundamental valuation approach to investing is that short-term volatility of stock pricing can provide additional incremental returns. By increasing (decreasing) exposure to equities during cyclical downturns (booms) when PE ratios are low (high), patient investors can receive excess returns.
What can be done?
To address short-termism at the fund manager level, we suggest investors:
- Commit funds for a defined period – which would be ideal, although unlikely to be practical;
- Extend performance measurement to five years (or more). Monthly, quarterly and yearly performance figures are not a sensible basis on which to make decisions on whether an active fund manager has performed well;
- Pay annual bonuses on the basis of rolling medium to long-term performance; and
- Require portfolio managers to make meaningful investments in their own funds.
Note: This is an abridged extract from the paper that won the Editor’s Pick Award for best Due Diligence Forum Research Paper at the 2011 PortfolioConstruction Forum Conference. The full paper is available at www.PortfolioConstruction.com.au.
Mark Arnold is chief investment officer and Jason Orthman is a portfolio manager/analyst at Hyperion Asset Management.
