Comments on The Ebb and Flow of Finances
A five year rule is no more than a rule of thumb and a dangerous one if inappropriately used: it comes from a general assumption that holding equities over a five year period will allow one to recover declines caused by recessions and or other significant economic risks.
I remember analysing the time frame of these risks back in the late 1980s and I would agree, that a five year rule is a vaguely reasonable (fuzzy) benchmark under normal conditions (a period of long term growth interspersed with recessions brought about, inter alia, by “normal events” such as monetary tightening due to short term excess demand as the business cycle matures and frays round the edges).
However, the 5 year rule does not look at total return/opportunity cost issues, and the real time frame is the time it would take for the total return on equities to be beat the total return on lower risk assets. Looking at historical risk periods, say the UK market in the 1970s, it took some 8 years from peak to peak for total equity returns to beat a lower risk alternative – transaction and management costs and relative performance issues if you hold a mutual fund would extend the time frame. There have been far longer periods of time in which the total return on equities failed to beat the total return on bonds etc.
However, the 5 year rule is not a constant (it varies): at the time the last major bull market peaked in 2000, the 5 year rule had extended beyond 10 years (I was looking at 12+) in many markets (ex ante) and much longer ex ante,(15 years +++) for the total return on equities to beat the total return on a lower risk asset.
Additionally, the 5 year rule does not really work when a market only brielfy regains it previous nominal high and then declines, because of course, the point relates to the ability to safely draw on equity capital beyond a certain time horizon. The 5 year rule (at peak valuations) has not applied in reality for a very long time (although clearly it is a winner at market troughs), and it must always be assessed relative to a wide range of risks.
When looking at modelling risk and return over time you need to assess what level of significant market and economic risk you would like to be able to cope with, and indeed one that reflects market and economic risks of the time: this is then the time frame over which you need to be able to meet income and capital needs without having to realise equities in the event of a decline.
A simple 5 year rule can be a dangerous rule, especially if you are having to draw on capital and have failed to build into your modelling (aside from the significant market and economic risks that should be a staple of risk modelling) a number of other risks that affect the ability of assets to meet financial needs over time.
I think the main issue for long term investors is not that they need to be able to take a 5 year time frame (many clearly have been invested for much longer), but that for many they have been subjected to derisory real returns and excessive risks for what is becoming a very long time indeed, all the while being told to a) calm down and b) that events “are normal”, which they are clearly not.
Normal risks are those risks which are meant to be covered by risk premiums built into asset prices. If asset prices are too high then the risk premium is too low, meaning that assets are priced for minimal risk going forward, meaning that risk which is not correctly priced by the asset is not really a normal risk.
One of the issues at the moment is that there is great uncertainty over the risks going forward: we do not know to what extent global debt is going to hamper growth or how much will need to be written off; these are all factors which we need to know to work out a risk premium. Historically, in a stable economic environment you can more or less assume that your capital stock will remain relatively stable and be added to at a certain rate and that growth will proceed within certain bands: at the moment, if we continue to delever, capital stock will decline and so will growth, meaning the ability to work out a risk premium is subject to a wide margin of error.
In other words, the time frame of risk may need to be extended as uncertainty over return, in an unstable environment, increases, irrespective of how far markets have already fallen.