The Struggle to Invest Rationally


R.B. Matthews and Doug McCutcheon

Richard Thaler won this year’s Nobel Prize in Economics. This is the third time in the past 15 years that the prize has gone to a behavioural economist. Clearly, our understanding of human behaviour is evolving rapidly.

Because natural selection happens over millions of years, we are all hard-wired to live in a world that is not the world we live in today. We are programmed to think and behave in certain ways. When it comes to investing, the result of such behaviour can be shockingly expensive.

When investing, our cognitive biases include over-valuing what we own, under-estimating the role of randomness, focusing too much on what has happened recently and allowing ourselves to be far more affected by a loss than by a gain of the same amount. We are also over-confident in our ability to predict the future. This over-confidence includes incorrectly believing that, in the past, we accurately predicted what was going to happen.

Today we live in an age when technology can assist us in various walks of life. This begs the question: if human beings consistently fall prey to biases when investing, why not program a computer to invest for optimal results? In fact, this is exactly what a number of investment firms (including Richard Thaler’s firm) are doing, under rubrics such as factor investing and smart beta investing.

Higher before-tax returns can be achieved by investing in the shares of companies with certain measurable attributes, such as a low valuation and a history of steady growth in earnings. Investment firms use computers to monitor these factors and automatically adjust their stock portfolios as the factors shift. Studies have concluded that these factor-based methodologies are, over the long run, likely to produce above average before-tax returns.

However, while this is useful for pension funds and other tax-exempt institutions, it is less helpful for individuals who live in an after-tax world. Taxable investors, when exposed to a high-turnover strategy, pay far more capital gains tax than necessary.

In addition to the behavioural biases traditionally studied by psychologists, we have noticed that investors find it almost impossible to think about investing and tax planning at the same time – another costly cognitive bias. Tax creates a strong headwind to long term wealth accumulation. And, while before-tax investment returns are only probable, the cost of tax is certain.

How, then, might we devise a system to mitigate our cognitive biases with a view to producing the highest possible after-tax return from stocks over the next, say, ten years? Even if you don’t have the option or desire to use a computer and a complex set of algorithms to invest your capital, here is a basic set of rules that should result, over time, in a superior after-tax result:

• Don’t be over-diversified. Twenty to 25 stocks are more than enough to provide sufficient diversification, but not so much that your portfolio starts to mirror the market.
• Select companies with low levels of debt, a history of steady growth in earnings and a below average ratio of stock price to earnings per share.
• For the portion of your capital allocated to stocks, remain fully-invested at all times. Recognize that timing the market does not work. What works is time in the market. (Remaining fully invested will be a hard rule for most people to follow, due to a combination of over-confidence in their ability to predict the future and risk aversion.)
• Plan to hold your companies for the long run, thereby deferring the realization of capital gains tax.
• Sell any stock in a taxable account that has fallen by, say, 20%, in order to trigger the capital loss for tax purposes. If you still think the stock will produce a better than average return, buy it back after waiting 31 days (to avoid the tax rules related to superficial losses).
• If you have both tax-sheltered and taxable accounts, allocate your stocks in order to minimize the tax you pay on dividends, a guaranteed way to increase your after-tax returns. For example, to the extent possible, put Canadian dividend paying stocks in your TFSA, U.S. dividend paying stocks in your RRSP and non-dividend paying stocks in your taxable account.
If you follow these rules, you will have eliminated many of the most expensive biases that cripple investors. With or without a computer, your long term after-tax return will be well above average.

Published in the Globe and Mail

Tags: