Noise: A Flaw in human Judgement is a bestseller and deservedly so. The book by Daniel Kahneman, Olivier Sibony and Cass Sunstein offers fresh insights to investing.
Although the book only touches on investing, the practical wisdom contained within is universal. It covers two types of error—primarily noise but also bias.
For the purposes of the book, noise is the differences between forecasts. Bias is the tendency for most forecasts to be consistently too high or too low versus the actual result.
For example, I brought up studies of the top economists’ forecast of the 10-year Treasury bond to Olivier Sibony. He explained that the top economists do not agree, and that variability is noise. Yet those studies also show they have, on average, consistently forecasted the rates too high. That is bias.
Relying on these economic forecasts would result in underperformance by staying on the short end of bonds, expecting rates to increase and bond ETFs to decline in value. Until this year, rates tended to decrease over a nearly 40-year period.
Consider two stock market forecasters, and decide which one you think is more credible:
Dr. Jones is very articulate, with sound and compelling logic and a high degree of confidence. In a panel of experts, she stands out as the expert’s expert.
Dr. Smith is far less sure of herself. She gives ranges of numbers and keeps talking about where her forecasts could be wrong. In a panel of experts, she is not the standout star.
It turns out that, in most fields, forecasters are barely more accurate than “dart-throwing chimpanzees”. But “superforecasters” (ones with better track records) are in perpetual beta mode (not a final version; always improving) by their open-mindedness in paying more attention to people who disagree with them than agree.
You are more likely to see the Dr. Joneses of the world on TV sounding confident about the top stocks or ETFs. The Dr. Smiths of the world, however, have a much better track record of forecasting. Your intuition may make you more likely to invest according to Dr. Jones, but no matter who you rely on, use the internet to look at their past track record. Few people actually do.
Two key findings
When it comes to investing – two key findings are relevant:
Mechanical predictions (models) are superior to those made by people.
Aggregating forecasts reduces noise (but not necessarily bias).
Applying the first finding can be seen in market events during a horrible 2020 in the pandemic, with surging unemployment, plunging GDP, social unrest and a dysfunctional political climate.
A robo-adviser had algorithms it followed and should have bought stocks during a 35% plunge in 33 days ending March 23, and reaped the rewards of a stock market that gained over 21% for the brutal year, using the total return of the Vanguard Total Stock Index fund.
Human forecasters had to consider countless complex health and economic data to analyse and likely did far worse.
The second finding is to view the stock and bond market as the ultimate in aggregating forecasts. The stock market is the aggregate forecast of millions of market participants of the discounted future cash flows of over 10,000 publicly held companies—not just the few talking heads on TV.
So a total US stock index ETF – such as the iShares Core S&P Total US Stock Market ETF (ITOT) – and a global index ETF – such as the iShares Core MSCI World UCITS ETF (SWDA) or the Vanguard FTSE All-World UCITS ETF (VWRL) – is a simple way to benefit from the aggregate of these forecasts, not to mention the benefit of low fees and high tax efficiency.
It takes into account the aggregate forecasts of the discounted cash flows of companies, industries, style, etc. Funds like smart beta ETFs and innovation ETFs are unlikely to both be right.
What about bonds?
I have already mentioned that the top economists had, in the aggregate, a costly bias to overestimate interest rates. But, like the stock market, there are millions of participants in the bond market rather than the 50-60 top economists.
Remember that if most of us knew the interest rate on the 10-year Treasury bond was going to increase, we would bid less in a Treasury auction to achieve that higher rate, so interest rates would have already gone up. That means a better forecast of that bond rate in the next year would have been the current interest rate of that bond, which represents the forecasts of millions of bond holders.
High quality bond index funds – such as the iShares US Aggregate Bond UCITS ETF (SUAG) and the Vanguard Global Aggregate Bond UCITS ETF (VAPG) – represent the aggregate votes on investment-grade, taxable fixed-rate bonds.
Stick to simple rules, like rebalancing the asset allocation to take into account your need and willingness to take risk. The noise that distracts from these simple rules is compelling. Though it will not be easy, investors can learn to reduce noise and bias.
Allan Roth is founder of Wealth Logic. He is required by law to note that his columns are not meant as specific investment advice.
This story was originally published on ETF.com
