Benjamin Graham famously stated that “In the short run the market is a voting machine, but in the long run, it is a weighing machine.” It is critical to keep this in mind as we discuss market valuations. In the short term, prices are set largely via reflexive forces related to informational cascades, herding, and positive feedback dynamics – so-called “animal spirits”. Markets that have gone up recently are more likely to continue going higher, regardless of valuations, as investors are primarily motivated by the fear of missing out on gains, and lagging their peer group. These dynamics manifest in the so-called “momentum” phenomenon, which has been observed since the dawn of markets.

However, high valuations set up a condition of ‘criticality’, such that as valuations become more stretched, even small shocks to the system can trigger a change of state. Think of a pile of sand. When the pile is small, pouring more sand onto the pile simply creates a larger pile. At some point the pile grows tall and narrow. As more sand falls on the peak, it continues to grow, but eventually a large portion of the structure falls off, collapsing the pile. It can’t be known in advance when the structure will break. But we can observe that the structure is becoming more and more fragile.

Current valuations imply that the market is approaching a state of criticality. So long as there are no major shocks, the market can press higher, perhaps substantially so. But the higher stocks go in the short term, the more vulnerable they become to a major event. This event cannot – to our knowledge – be reliably predicted in advance. However, there are methods – that have nothing to do with valuation analysis – that may be useful in identifying a change of state shortly after the fact, and that may be used to limit losses. We cover such techniques elsewhere, as they are beyond the scope of this article.

To our mind, while valuation analysis is not an effective tool for market timing, it is still extremely useful, primarily for estimating future long-term market returns. When you smooth away all the market noise, the more you pay for a dollar of earnings or balance sheet today, the less you should earn on your investment in the future. Valuation analysis can help quantify this relationship, to help capital allocators make sense of all the available investment options.

Herbert Stein is credited with the tautology that “If something cannot go on forever, it will stop.” Other colloquialisms like “No tree can grow to the sky,” express a similar sentiment. The formal name of this effect is “reversion to the mean”, which is simply the principle that most systems have a reasonable average level, around which the system fluctuates with some random error. This is an extremely useful principle because it allows us to identify when a system has strayed too far from its normative value, and quantify a likely trajectory.

Most analysts take the long-term average valuation level for their equilibrium value. This method is parsimonious and defensible, but it makes the assumption that the variables that factor into valuations do not change through time. While this may indeed be the case, for several reasons, we have cause to question this assumption.

Should Valuations Expand Over Time?

Notwithstanding myriad fundamental factors that influence equilibrium market valuations, the fact is that time itself should lead to a higher equilibrium. Consider that, as markets progress through time, they present investors with more data. Investors who take the time to study this data may become more confident about the market’s true character. Participants might feel they have a better grasp of the distribution of risk and expected payoffs, and therefore require a lower return to deploy capital, leading to higher average prices over time.