When I started out in banking, I was based in a dealing room advising traders on potential positions to take. The positions were focused and generally very short-term in nature. Therefore, risk management was not a ‘nice-to-have’, it was vital to job security. When entering a trade, a stop-loss – a level at which the position taken would be unwound if it was losing money – was a must.
It was against this backdrop that a former colleague quipped that ‘stop-losses are for wimps’. He was of course referring to certain stocks in his portfolio which had fallen dramatically – he was probably justifying to himself why he should keep it! However, it raises an interesting question: Should we employ stop-losses when we invest?
While many people will be very passionate about this topic, as with most things in life, context is key. If you think about it, the existence of the stop-loss is a hedge against the fact that nobody knows what is going to happen and therefore you need to build in a circuit-breaker to avoid the behavioural biases that come with a loss-making position – the ostrich, or ‘head in the sand’, syndrome.
What does it mean for an average investor?
So how does this translate for an average investor? I would argue that there are two dimensions to consider: the nature of investments being discussed and your time horizon. Let’s take each in turn.
I have a much greater conviction level that a diversified ‘foundation’ allocation (which includes exposure to different asset classes such as stocks, bonds, gold and private assets), or even a diversified equity portfolio, is more likely to rise over a given period of time than any individual stock.
The reason is simple. Different asset classes have different drivers and hence are usually uncorrelated in their moves. Therefore, just combining them into a portfolio smooths out the bumps and increases the probability of positive returns for the portfolio as a whole.
When it comes to the stock market, there are also different drivers that determine equity market or individual stock returns, but let’s simplify them into broad market drivers and idiosyncratic drivers. If the economy enters a recession, then most stocks will fall sharply in value. However, a company’s failed product launch will largely hit one stock, and those of a few of its suppliers. Its impact on the overall market will be much less severe.
This brings me to Principle Number 1: The broader the investment, the less likely a stop-loss is warranted and that a buy on dips approach makes sense.
Probability of rebound
A very good friend of mine recently questioned how applicable this ‘buy on dips’ approach was to stock markets outside of the US. So, we ran the numbers for some major global and Asian markets in terms of probability of positive returns over different time horizons and the potential size of returns for investors. The results are pretty interesting.
First, the historical probability of positive equity market returns across any given 12-month period, at around a two-thirds probability, is generally similar across major global or Asian indices – China is an outlier at just 55%. If you extend your time horizon to 5 years, this probability generally increases to around 80% - the outliers are Japan’s TOPIX (66%) and India’s Sensex (92%).
Second, we looked at what has happened after a 10% or 15% market pullback. Focusing on the 1-year time horizon, we can break the countries into 3 groups:
1) either the probability of positive returns or size of average return or both have increased significantly after a market sell-off. Markets that fall into this group include the US, Germany, UK, India.
2) there is no material change in either variable. This includes Hong Kong, Malaysia and Korea. Once you lengthen your time horizon to 5 years though, they all move into the first group.
3) the probability of positive returns or their average quantum actually declines after a sell-off. Japan and onshore China markets fit into this group. On a 5-year time horizon, China also moves into the first group but, interestingly, Japan stays firmly in group 3.
Hence the conclusion is: outside of Japan, the ‘buy on dips’ mantra has made sense, especially when held by long-term investors.
The above analysis highlights the importance of the last factor: time horizon. The dealing room environment generally takes narrow exposures over a very short time horizon. Thus, stop-losses are crucial. However, I believe the message for the average investor is the reverse as long as they are focused on a diversified foundation allocation with a long-term focus.
Structural thematic investments also potentially fall into this category. The longer your time horizon, the more likely your investment is to generate positive returns as long as the structural fundamentals remain supportive. Here, again, instead of a stop-loss being appropriate, market declines offer an opportunity to add to longer-term thematic positions given the likelihood that declines will prove temporary.
Principle number 2: The longer your time horizon, the less desirable a stop-loss is, especially for diversified allocations or long-term structural themes.
Thus, for investors who are trying to trade the market and pick stocks, we believe a strict risk management framework including the use of stop-losses is absolutely critical to returns. However, we believe the majority of investors would be much better served by building a foundation allocation with a ‘buy on dips’ approach. Investors can systematise this ‘buy on dips’ approach through regular portfolio rebalancing – say at least once or twice a year and especially after major market dislocations to bring their allocations back to their desired risk tolerance. Such rebalancing is akin to an investor systematically “buying low and selling high” – a win-win proposition. For these investors, stop-losses are likely to get in the way of wealth accumulation.