Search
  • Stuart Fowler

Approaching old peaks: so what?


Stock markets around the world are closing in on the index levels achieved before the last bull market turned into a bear market. With hindsight, we can view the old peaks as bubble territory, inflated by foolish notions about technology. Six years later, it is tempting to think the same levels must again be dangerous and to view any near-term weakness as the end of another bull run.

In real terms, adjusted for inflation, and allowing for the long term trend in ‘total returns’ (including income) for major equity markets, today’s levels are mostly not far from half the equivalent peak. That doesn’t tell us anything about what will happen next. But it does mean the downside risks are no greater than ‘normal’, and no greater than the upside potential, whereas at high deviations above trend the odds are stacked against you, with much greater downside risks than upside potential.

Real total return: the bottom line

Investors don’t have many certainties to hang their hat on when it comes to equity investing. About the closest is that an entire market (such as represented by a broad index – or index-tracking fund) will generate a ‘trend’ real (inflation-adjusted) total return (income and capital growth) of about 6-7% pa.

In the chart we show the path of real total returns for the UK. The trend fitted for the entire period (so including hindsight) is shown in yellow and the trend fitted to data known only at the time (hindsight-free) is shown in pink. The data runs from 1900 and the full-period trend is 6.2% pa.

Comparable trends and data lengths for other markets with similar data evidence are:

– Europe ex UK: 6.9% (1970)

– USA: 6.9% (1926)

– Japan: 5.6% (1957).

The deviations around this trend are also quite similar for different markets. Clearly, they are large, are themselves serially correlated (they also trend) and may not correct or reverse quickly. But the reversion to a sustainable mean return when measured over long periods is too prevalent over history and between different markets to be an accident. There is something else going on.

Professional investors and academics can argue over what it is that explains either the deviations or the correction process but for all practical purposes any investor knowing the history can accept it without a highly-developed theoretical proposition or proof.

Market conditions matter

Accepting reversion to a sustainable mean real growth in return is not the same as saying that if you wait long enough you will earn that return. This is a widely misunderstood reading of history.

Those who bought UK equities in the late stages of the bear market of 1973/4 enjoyed a much higher real rate of return when measured over the next two decades, even without the benefit of the later technology boom. Similarly, those who bought Japan at its peak in 1989 were condemned to much lower than trend returns for two decades and are still waiting. Clearly, buying at times of high deviation from the sustainable trend, high or low, makes a big difference to the long-term growth rate enjoyed. Market conditions matter a lot.

In the chart below we show how No Monkey Business Limited plots the ‘real valuations’: as a ratio of the actual level at the time to a trend for the whole period.

In each case, a ratio of zero (these are logarithmic, to show the proportionality better) is equivalent to current market conditions being exactly in line with trend, so we need make no adjustment to the normal trend return if making an investment at such a point and calculating the expected real return. The extreme deviations in this chart are roughly equivalent to twice overvalued and half undervalued. We can note that the 1999 technology-inflated boom ranks for the UK, US and Europe as extreme. Indeed, market conditions were not dissimilar to the Japanese bull-market peak a decade earlier. Note that sometimes the deviations are similar but at others one or two markets may behave quite differently (or different in degree) from the others.

Comparable trends and data lengths for other markets with similar data evidence are:

  • Europe ex UK: 6.9% (1970)

  • USA: 6.9% (1926)

  • Japan: 5.6% (1957)

The deviations around this trend are also quite similar for different markets. Clearly, they are large, are themselves serially correlated (they also trend) and may not correct or reverse quickly. But the reversion to a sustainable mean return when measured over long periods is too prevalent over history and between different markets to be an accident. There is something else going on.

Professional investors and academics can argue over what it is that explains either the deviations or the correction process but for all practical purposes any investor knowing the history can accept it without a highly-developed theoretical proposition or proof.

Market conditions matter

Accepting reversion to a sustainable mean real growth in return is not the same as saying that if you wait long enough you will earn that return. This is a widely misunderstood reading of history.

Those who bought UK equities in the late stages of the bear market of 1973/4 enjoyed a much higher real rate of return when measured over the next two decades, even without the benefit of the later technology boom. Similarly, those who bought Japan at its peak in 1989 were condemned to much lower than trend returns for two decades and are still waiting. Clearly, buying at times of high deviation from the sustainable trend, high or low, makes a big difference to the long-term growth rate enjoyed. Market conditions matter a lot.

In the chart below we show how No Monkey Business Limited plots the ‘real valuations’: as a ratio of the actual level at the time to a trend for the whole period:

In each case, a ratio of zero (these are logarithmic, to show the proportionality better) is equivalent to current market conditions being exactly in line with trend, so we need make no adjustment to the normal trend return if making an investment at such a point and calculating the expected real return. The extreme deviations in this chart are roughly equivalent to twice overvalued and half undervalued. We can note that the 1999 technology-inflated boom ranks for the UK, US and Europe as extreme. Indeed, market conditions were not dissimilar to the Japanese bull-market peak a decade earlier. Note that sometimes the deviations are similar but at others one or two markets may behave quite differently (or different in degree) from the others.

Knowing that today’s market conditions are merely close to average tells us what we need to know about long-period projections. The mean of those projections will be close to 6% pa (before expenses), not 3% or 12%. The rate actually earned might still be as low as 3% or as high as 12% but the chance of either extreme being reached is less than 1 in 20 when ‘normal’, not 1 in 2 when extreme.

Knowing what long holding-period projections to make does not help us very much in the short term, such as 1 or 2 years. Over these ‘decision time frames’ markets are essentially a random walk. History tells us what risks to be aware of (what can happen) but not what will happen, or what will happen next.

You can diversify your stock exposure, as in an index, and you can diversify your geographical exposure, as in some mix of all four major markets, but beyond those actions you cannot cut your risk without actually limiting your esposure – by investing less in equities as a whole or by insuring your downside risks.

When risk choices are informed by probabilities, such as those generated by a model like this, they can appear quite simple – even if it is hard to accept the lessons of historical real returns and variances.

#equities #models

0 views

© 2016-2020 Fowler Drew Limited

  • Black Facebook Icon
  • Black Twitter Icon
  • Black LinkedIn Icon

Authorised and regulated by the

Financial Conduct Authority