Long-term investing is about objectives not risk

Investors typically believe that stocks are less risky in the long run.

Some lean on this belief to extend their time horizon when faced with unrealised losses.

As one investor argued, stocks mean-revert and, hence, carry lower risk over the long term. Does mean reversion justify long-term holding of stocks?

This article discusses mean reversion and explains why it is counter-intuitive to hold mean-reverting assets for the long term.

It then argues that investors hold assets for the long-term because they are intuitively aware of asymmetric returns-compounding effect.

It also shows how the core-satellite framework helps in moderating these factors, encouraging holding stocks for the longer term.

Mean reversion

The argument that stocks are less risky in the long run is puzzling. For, if stocks are risky in the short run, where does the risk disappear in the long run? One argument could be that investors are able to predict variation in stock price over the longer horizon. This type of variation that reduces long-horizon risk is called mean reversion.

This means that a short-term downtrend in asset prices will be likely followed by an uptrend, as prices correct to their long-run average. Likewise, a short-term uptrend in prices will lead to lower returns expectations, causing price correction in the future. All of this means that price risk reduces over the longer horizon.

Now, the basic principle of financial market is the risk-return trade-off.

This means that the investors can only expect returns that are commensurate with the associated risks.

If mean reversion were indeed true, the long-run price risk for stocks will be low. And that would translate into lower long-run returns.

So, why then do investors extend their time horizon? The answer, perhaps, rests in the fact that investors are intuitively aware of asymmetric returns-compounding effect.

Asymmetric compounding

Suppose an investor buys a stock for Rs 10,000. Assume that the stock can only take two states — it can either rise 50 per cent or decline 50 per cent. If the stock declines in the first year by 50 per cent to Rs 5,000, it has to move up 100 per cent in the next year to recover the initial capital. Can it?

Even if the stock were currently way below its long-run average, it may take a long while for it to claw back the initial loss. Suppose instead the stock had risen 50 per cent in the first year to Rs 15,000.

It has to lose only 33 per cent to fall back to its initial investment.

The above example demonstrates a fact about returns-compounding — it takes a lot of effort to recover capital than it is to lose it!

Returns-compounding can, hence, have a devastating effect on the portfolio.

Investors intuitively understand this effect.

That is why they conveniently convert their short-term holdings into long-term investments when there is a large loss in the portfolio.

The problem, however, is that mean-reversion questions long-horizon investment while asymmetric compounding encourages such investments. What then should investors do?

Conclusion

Investments should be driven by liability structures such as buying a house or protecting pre-retirement lifestyle, post-retirement.

Mean reversion and asymmetric returns-compounding would then be drivers for the strategic asset allocation decision during the portfolio construction phase and later, for portfolio rebalancing.

The debate on whether stocks are less risky in the long run arises from the need to shift assets from short to long horizon.

The core-satellite portfolio overcomes this need for horizon-shifting by separating short and long-horizon exposures.

The core retirement portfolio, for instance, is passively-constructed with a horizon to match the retirement date while the satellite is structured to provide excess returns through continual rebalancing.

Such structures do not settle the debate on whether stocks are less risky in the long run; they help investors' side-step the issue and construct meaningful portfolios.