We review the last few decades of evidence-based research on how markets work and the implications for how individual investors manage their total wealth.
UBS, Switzerland’s largest bank, thinks new weight loss drugs will be “the biggest drug ever” with a market exceeding $150 billion by 2031 (The Economist, March 2023). An early entrant, Novo Nordisk, with the drug Semaglutide, has already seen its share price double in two years. Elon Musk is a self-declared fan. In trials, recipients lost 20% of their weight on average. The market is huge (obviously): the World Obesity Federation estimates that half of everyone over five will be overweight or obese by 2035. Semaglutide must be taken for life to be effective at a monthly cost of $1300.
Should investors be salivating over obese profits? Is Novo Nordisk overvalued or undervalued at the current price?
To decide whether to buy Novo Nordisk (or any security) we need a pricing model that generates a price we can compare with the current market price. If the model predicts a higher price, then we would buy the security. Conversely, we would be a seller if we owned the stock and the model predicted a lower price. This makes sense if the market price was based on just another model and had no more weight than our own estimates – as if Mr. Market was just a guy in a bar with his own opinions. But Mr. Market has a superpower: he is the collective wisdom of all market participants, and the market price reflects the collective wisdom of everyone’s analysis. To gain an edge over Mr. Market we need to:
- have relevant and public information that hasn’t yet been incorporated into the market price or
- have a better asset pricing model.
This first issue relates to whether markets are informationally efficient – how quickly does the market price adjust to new information? The challenge is to disentangle this proposition from the asset pricing model: if expected returns differ from observed actual returns is this because of an inefficient market or a bad pricing model?
This was the challenge taken up by Eugene Fama in the 1970s and his research culminated in him being awarded the 2013 Nobel prize in economic sciences.
Fama and his collaborators started with an initial hypothesis that says that security prices should equal the expected discounted cashflows. Each of the words in bold requires explanation.
Cashflows. If we have invested in a company, we want payment for the use of our money in addition to the return of our capital. For a stock, the payment could be in the form of dividends or capital appreciation when eventually sold.
Expected. When thinking about the future we recognize that there are a range of possible outcomes. We assign a probability to each and sum the weighted components. For example, suppose we think the cashflow to all shareholders has a 25% chance of being $20 million and a 75% of the being $40 million. The expected cashflow is $35 million (0.25*20+ 0.75*40).
Discounted. A dollar today is worth more than a dollar in the future because today’s dollar can be invested at least at the risk-free rate. If I invest $1 for a 3% return without any risk, then I need at least $1.03 a year from now from my shares. If I could get a 5% return risk free and the cashflow from selling my stock in a year is only expected to be $1.03 then when I discount the future $1.03 to the present it is only worth 98 cents (1.03/1.05), indicating that the proposed investment is not attractive. Riskier stocks have a higher discount rate and a lower estimated security price from the asset pricing model.
Over short periods, the discount rate is not expected to change much so early studies testing market efficiency looked at events like stock splits or earnings announcements and merger announcements which occurred over a few days or months.
The conclusion from these studies was summarized in Fama’s Nobel Prize lecture, “…early event studies generally confirm that the adjustment of stock prices to events is quick and complete”[1]. Exactly the prediction of market efficiency. Subsequent studies have led to a similar conclusion.
The conclusion that markets are informationally efficient means that market prices incorporate all available information. The implications for investors are profound:
- Market prices represent the best estimate at any time and selling one stock for another stock (stock picking) has no expectation of an excess return above market returns. We choose our words carefully here. It may be the case that the stock you buy today increases in value tomorrow – there is a lot of unpredictable noise in stock prices. The efficient market model states that if you earn above market returns this is due to luck not skill. Like rolling a die or tossing a coin, in the short term you are gambling (i.e. playing a game of chance) and over the long term you do not outperform rational expectations.
- For markets to remain efficient there needs to be some stock pickers. Some of these stock pickers will make bad decisions which will be balanced by other smarter stock pickers on the other side of the trade. In aggregate these active traders perform a public service by driving prices to equilibrium but, logically, they cannot outperform the market because, combined with passive investors, they are the market2.
If markets were not efficient, then we would expect to see persistent outperformance by smart investment managers who identify mispriced securities through stock selection or market timing. Standard and Poor’s (S&P) has been comparing the performance of fund managers with the relevant market index for years. S&P split the funds into 7 different categories including Canadian Equity Funds, Canadian Dividend and Income funds, U.S. Equity funds and Global Equity funds. For each category they compare the aggregate fund performance with the relevant market index for their fund category. Their most recent (March 2023) report3 reviews performance to the end of 2022.
In 2022, a little over half the funds underperformed which is consistent with short term performance being unpredictable and close to a coin toss. As the measurement period lengthened the results deteriorated so, depending on the fund category, 72% to 98% funds underperformed their index after ten years. In other words, the longer your investment horizon, the less active managers are likely to outperform the market.
Other studies have shown that fund managers that have done well in one year have no better chance than their peers to do well in subsequent years. The Persistence Scorecard for the U.S.4 notes for example “Of the actively managed equity funds whose 12-month performance placed them in the top quartile of their respective category as of June 2020, not a single fund maintained its top quartile performance over the next two 12 month-intervals (their emphasis)”. The report concludes “over a five year horizon it was statistically near impossible to find persistent outperformance”
Equipped with these insights into the information efficiency of markets and the inability of professional fund managers to sustain market outperformance, we can now revisit the weight drug opportunity at the beginning of this article.
We should assume that the current Novo Nordisk price reflects the expected discounted cashflows and is neither overvalued nor undervalued. If there is news tomorrow that the obesity drug has previously unknown side-effects, for example, then we expect the price will have taken this into account by the time we get to hear about it. Mr. Market is a pricing machine and not easily outwitted.
If we can’t make money from identifying mispricing opportunities (aka “beating the market”), where do investment returns come?
We will explore this in a future article.