A system that creates unforced errors
When we set up Market Thinking in 2019, a key part of the motivation was to have the freedom to do things differently now that we were on ‘the outside’. After more than 35 years of working for large financial institutions we wanted to have the freedom to not do things in the conventional way, mainly because as ‘insiders’ we understood (some of ) the downsides of the system and in the words of the late Charlie Munger
“It is remarkable how much long-term advantage people like us have gotten by trying to be consistently not stupid, instead of trying to be very intelligent”
Charlie Munger
Charlie of course was extremely intelligent, but his point is nevertheless well made, a vast number of highly intelligent people in the world of investing consistently do ‘stupid’ things because they operate in a system that forces them to do so. Another great Charlie Munger quote is “Show me the incentives and I will show you the behaviour” and this is the point we wished to make when setting up Market Thinking; if we can now step outside of the system and if our incentives are now lined up with the underlying investor rather than the Investment Management firm that pays the bills, what might we do differently?
To bring it up to date and put another way, if we were to take the proverbial clean sheet of paper and write on it our New Year Resolutions for better investing in 2024, what might they be? Perhaps this:
- Construct a portfolio of quality stocks that we believe offer the prospect of long term compound growth.
- Look to achieve some diversification and recognise that being in a good boat in a fast part of the stream is better than being in a great boat in a slow part
- Invest with conviction, cash is a hurdle to be beaten, but also a key risk management tool. There are times when it is right to be heavily in cash.
And yet, most professional investors are prevented from doing any of the above, and unable to follow simple ‘Rules for not being stupid’. They (previously we) have to construct portfolios with an obsessive reference to the ‘benchmark’, since that is how the management of fund management firms judge them (show me the incentive) and the ‘risk management’ largely takes the form of tracking how much benchmark risk they take and how much volatility there is in the portfolio.
The latter tends to take the form of Value at Risk or VaR models, which are essentially designed to measure the risk exposure of trading desks in investment banks. Applying it to long term investments makes no sense, but as they say, what can be measured will be managed and so it is. The former, meanwhile, leads to an obsession with ‘alpha’, how much a manager can beat a benchmark on a monthly, weekly or even daily basis. Much of this is noise but nevertheless creates an unhealthy culture “I can’t sell that product as it underperforms the index” leading to managers taking, often hidden, risk in order to ‘outperform’. Thus, while official risk management looks at short term volatility and tracking error, the managers will often take significant stock specific risk that is not properly picked up. Sometimes it will be exposure to highly or even moderately illiquid stocks, as a virtuous circle builds whereby a ‘winning’ manager attracts inflows and deploys them into existing large holdings where they are not only a significant part of the manager’s portfolio, but that the manager is a significant part of the overall shareholder base. Traders understand this of course and will actively ‘job’ these mutual fund flows, accelerating the effect of an upward squeeze. More often than not these positions unwind (sometimes painfully) with the same mechanism and the winners suddenly lose their crown. Sometimes the stocks themselves blow up; for example when Wirecard blew up, several ‘highly successful’ managers had it as more than 10% of their portfolios. Crowded trades become crowded shorts as the traders exploit the system that enables/forces clever people to do dumb things.
Generally, though, most active managers will have a number of genuine conviction bets in their portfolio, but concede that they also have a large amount of ‘ballast’ to keep their ‘tracking error ‘acceptable. This means having a global or US portfolio that will have to own Apple, Microsoft and Nvidia regardless of any view on their actual prospects. It will also mean buying and selling them in effectively a passive manner. At its extreme it means having 10x as much in Apple compared to, say, Merck, which is nevertheless the 21st largest stock in the US market, which objectively makes no sense from a portfolio construction point of view, but is “just how it is”. All of this means that incentives create the situation whereby the client return becomes a function of the Asset Management Firm’s risk, a classic example of the Principal/Agent problem.
Secondly, the system forces sector and geographic allocation. Unless you are a country or sector specific fund, most managers will have an allocation to economic activity dictated in a large part by the existing market capitalisation of the markets. Thus back in the late 1980s, almost half of Global Equities were in Japan, that is now under 7%, while almost 60% are now in the US and, as we know, almost a third of that particular market consists of just seven stocks. Indeed if we look at the Vanguard World Equity ETF, we see that Merck is at number 26 in terms of size, with once again 1/10th the weighting of Apple, but that the only non US stocks above Merck are Samsung, ASML and TSMC. Is this really the best opportunity set? Or is it ‘just how it is’?
Third, the mutual fund ‘system’ generally only allows a maximum of 10% in cash, which is really a 3-5% buffer for meeting redemptions, cash calls etc with 5-7% for ‘risk management’. Thus even if the manager is absolutely convinced that the prospective risk return is highly negative, the best they can do is only have 90% of clients’ funds in the market - and most likely hugging the index. Historically, we have likened it to being convinced their is a major pile-up coming and instead of being able to pull over to the side of the road being told to trust that the airbags worked. Once again, the incentives are misaligned, rather than aiming to lose the least money when (not if) you get it wrong, the incentive is simply not lose more than the benchmark. Thus it was acceptable to be exposed to the Magnificent 7 in 2022 when they underperformed the equal weighted S&P500 index by 40% because you went down ‘in line’, but being exposed this year when they all bounced back was ‘great stock picking’.
The second source of Not being Stupid. Ourselves
If we can take away the unforced errors from the system, we can, in Charlie Munger’s words, hope to ‘gain some long term advantage’, but we believe that there is a second source of unforced errors we can also avoid, where we can gain an edge by ‘not being stupid’, which is to take the emotion out of our investing. This is the world of Behavioural finance; Loss aversion means we risk running our losers but cutting our winners, anchoring and recency bias risks making our decisions from an incomplete dataset, underwritten by confirmation bias preventing us from hearing any challenge to our view. Meanwhile herd instinct and self confirmation bias lead to bubbles forming - for example, ‘everyone’ loved Meta at Christmas 2021 at 360. hated it in December 2022 at 120, but having bought it in July at 320, they all love it again. But not as much as Nvidia, which they bought in August. When they also bought Bitcoin again.
Nobel Prize Winner Daniel Kahneman, who discussed many of these behaviours in his 2011 book Thinking Fast and Slow, also conceded that, despite having literally ‘written the book’ on the subject, he himself is still prone to these errors. As we all are. To overcome them, we need to build our own system of checks and balances as well as recognise that their existence allows for opportunity as well as threat. We know for example that Asset allocators will tend to buy things that are going up if they are ‘underweight’ and sell them if they are going down and they are overweight as the very system we are implicitly rejecting ourselves is still making them do the ‘stupid stuff’. (To repeat, it is not that the people are stupid, they are merely caught up in the system). Equally, we know that leveraged short term traders will tend to exaggerate herd instincts and create narratives to feed confirmation bias. As long term investors looking for compound returns these unforced errors from others need to be part of our information set.
Conclusion
With a Clean Sheet of Paper and a focus on Client Risk and Return, we would look to create a relatively (but not excessively) diverse portfolio of stocks weighted according to our conviction of potential compound risk adjusted return, rather than weighted by simple market cap and where risk allows for more than short term volatility. However, recognising that the rest of the markets are still forced to follow the ‘system’ is important, both in understanding the risks and opportunities due to probable behaviours of other market participants and in accepting that our way of doing things will not necessarily ‘fit’ with the rest of the ‘system’ either. In what ‘box’ would such a fund ‘fit’ for example? How would it get past the gatekeepers who have also spent 35 years inside the system? This is of course why the system persists. Incentives.
As a further and final twist, regular readers may recognise that we have been applying these disciplines in our model portfolios since 2019 and in our managed portfolios since 2020, but without going down to the stock level. Instead we prefer to eliminate another source of potential unforced error in the form of stock specific risk by using large and liquid ETFs to capture Global Factors and Themes.
For more details on the UCITS fund we have launched this year with ToscaFund, one of the UK’s leading alternative asset managers, eligible investors can visit the ToscaFund HK website here.
