This article is Part One of our ‘Investing – An Alternative Approach’ miniseries. The series aims to inform readers about how the Frame Long Short Australian Equity Fund generates returns for its investors.
When launching a new strategy, the first thing to consider is its purpose in the investment universe. Our objective (along with many other individual and institutional investors) is relatively simple – we aim to outperform a market index on a risk-adjusted return basis, providing a superior alternative to passive equity index investing or other actively managed funds.
In order to design such a strategy, we first need to consider the dynamics of the index we want to beat. Most readers will be familiar with the S&P/ASX200 index. It comprises the largest 200 publicly traded companies in Australia, weighted by their market capitalisation (subject to liquidity conditions). Simply, the index is just another trading model – it ‘buys’ stocks when their market capitalisation rises and ‘sells’ them when it falls.
This means the manager is already on the back foot when benchmarking to an index. We are effectively trying to beat a trading model that has no fees, commissions or slippage. Looking at the index this way, it becomes apparent a ‘repackaged’ S&P/ASX200 product with fees will not succeed. Some outside the box thinking is therefore in order.
We start by examining some of the index’s flaws. The most obvious problem with investing in indices that are market cap weighted is unintentional destruction of diversification. For instance, the top ten stocks in the S&P/ASX200 contribute fifty per cent of index weight, while the bottom quarter of the index constituents contribute approximately three per cent.* This means index returns are dominated by the performance of a small number of very large companies, while the remaining index constituents contribute very little, even if they ‘outperform’ the broader market.
Given this design flaw, what happens if we remove the index’s bias towards large companies? We take a portfolio of 50 equally weighted, randomly picked stocks within the S&P/ASX200 universe (using historical index constituents) and rebalance monthly. That is, we randomly select 50 new companies at the start of each month to invest in for the next month. Running 1000 Monte Carlo** simulations on this strategy for the 20 years ending 1 January 2021, we find the average return is 10.68%. The S&P/ASX200 returned 8.03% inclusive of dividends for the same period! This means by selecting 50 random stocks and allocating equally each month, you would have had a high probability of outperforming the market. In our trial, only the bottom 1.5% of simulations returned less than the market return of 8.03%.
While this simulation is very simple and doesn’t include any transaction costs, its relative success gives faith to the idea that outperforming a benchmark index over the long run is a real possibility. By approaching the market from a different point of view rather than ‘index tracking’, we may be able to develop statistically significant strategies that achieve our objectives of outperformance on a risk adjusted basis.
Part two of the miniseries will cover the reasons for investing systematically, including an investigation of why humans make bad investing decisions and the use of automation to limit emotion.
*As of August 23, 2021.
** Monte Carlo simulations generate a range of possible outcomes by running a strategy with a random element repeatedly. This allows an approximate probability distribution of the strategy to be constructed and analysed.