13 May 2019
By John Husselbee, Co-Fund Manager HC Verbatim Portfolio Growth Funds
New research from the US suggests that while fund managers often display clear skill when buying stocks, their track record on selling is much weaker and could potentially be damaging long-term returns.
The crux of the report – entitled Selling Fast and Buying Slow, produced by academics from Chicago School of Business, Carnegie Mellon University and MIT – is that buying and selling assets are driven by different psychological processes. Cutting through the science, buying decisions appear more forward looking and belief-based while selling is typically more backward looking.
A quote from one of the managers questioned sums this dichotomy up: “Buying is an investment decision, selling is something else.”
On the buying side, the process for an active long-only investor typically involves in-depth research, often meeting company management, painstaking financial analysis and ultimately establishing the opportunity cost of not investing. Every buy decision is seen as offering a chance to shine, to “find the next winner” as the report puts it, and data appears to show this process broadly bears fruit, with freshly purchased stocks tending to outperform.
When it comes to selling however, to quote from the report once again, decisions are often driven by “asymmetric allocation of cognitive resources”, which means much of the effort that goes into buying decisions appears to be jettisoned.
It is important to understand that what the research calls ‘properly considered selling’ – on the back of new information that affects a company’s share price – still tends to yield a positive result. On so-called ‘non-announcement’ days however, sell decisions actually underperform a completely random disposal strategy by around 2% a year.
Elsewhere, the report also offers up interesting behavioural nuggets when it comes to parting company with stocks: some managers report a process more akin to obliteration than selling, completely removing the name of the company from their investment universe. Many also avoid tracking performance of stocks after selling, shunning the opportunity cost part of the process inherent when it comes to buying.
How does all this chime with our own sell discipline and that of our underlying managers?
First off, as with any broad research – the US study covered close to 800 institutional portfolios with an average value of about $570 million between 2000 and 2016, and 4.4 million trades in total – the findings are interesting but inherently reductive, largely missing the pool of genuinely active managers in which most fund selectors fish.
As anyone aware of our process will know, one of our core beliefs is that while consistency of performance is ultimately impossible, consistency of approach is vital. We select funds that complement each other in terms of investment style so it is key managers have a record of maintaining that style and do not drift away from it: if they do, this is a reason for us to sell out.
Because consistency of style is such an important element for our managers, a strong sell discipline is a key part of that and I believe funds in our portfolios buck the ‘poor sellers’ trend. As many have such strict buying criteria, selling is a natural counterpoint to that: as soon as a company no longer has the attributes that made it attractive, it will typically be sold.
As a general rule, we tend to prefer managers running more concentrated portfolios in terms of number of holdings. This means they should have a greater understanding of each stock and the case for continued inclusion than possible for a large institutional fund with hundreds of companies and potentially a long tail of holdings.
Take Jupiter European, run by Alexander Darwall, as an example, which I have owned for the best part of two decades. Darwall recently announced he is to give up the portfolio later this year but the point remains valid.
At a recent catch up with the team, they reiterated their aim to hold 35 ‘special’ companies and as that naturally involves building a complex picture of each position, it should facilitate effective selling. To be clear, selling here is unlikely to be based on short-term newsflow but rather a fundamental shift in the initial buy argument.
Belying the idea managers pay little attention to how stocks perform post sale, the Jupiter team monitors this closely: again, ongoing interest in one of 35 positions is likely to linger more than a few sells among hundreds of holdings.
Since 2010, Jupiter European has sold out of 26 companies completely and of these, 19 (73%) went on to underperform the benchmark, a record considerably better than the ‘worse than random disposals’ espoused in the US research.
Adopting an amateur psychologist’s hat for a moment, it is easy to understand why buying is a more enjoyable stock market experience, unless selling to bank a large profit. But a strong style demands a strict selling strategy as well as for buying and it will continue to be something we want to see in all our underlying managers.
The value of investments and any income from them can go down as well as up and is not guaranteed. Your clients could get back less than they originally invested. Past performance is not a guide to future performance. The portfolios' investments are subject to normal fluctuations and other risks inherent when investing in securities. Verbatim Asset Management has taken due care and attention in preparing this document, which is solely for the use of professional advisers. Verbatim cannot be held responsible for any inaccuracies arising out of information detailed within and will not accept liability for any loss arising out of or in connection with its use. This article is for information only and should not be deemed as advice.