ADRIAN CLAYTON: Best payback is to build knowledge about businesses in which to invest
Findings of studies on the predictive efficacy of forecasters are grim
13 January 2025 - 05:00
byAdrian Clayton
Support our award-winning journalism. The Premium package (digital only) is R30 for the first month and thereafter you pay R129 p/m now ad-free for all subscribers.
Forecasters may tell you a great deal about the forecaster; they tell you nothing about the future. — Warren Buffett.
Typical for the start of a new year, a conveyor belt of leading economic and market clairvoyants (forecasters) prognosticate on everything from gin prices to stock market outcomes. Despite the endorphin release from listening to their prophecies, do they add value? Do forecasters have real predictive power, and can you make money from their advice?
Buffett is clearly a nonbeliever, but then he is not always right either. Academic literature covers the topic well. Various studies have focused on the predictive efficacy of forecasters. Unfortunately, the findings are rather grim.
Stephen McNees and John Ries, economists at the Federal Reserve of Boston, reviewed the performance of a group of leading forecasting organisations based on their quarterly forecasts of economic variables between 1971 and 1983.
This study showed that the best forecaster in one year did not repeat their success in the next. Forecasters demonstrated an inability to forecast across a range of variables, and even when a specific forecaster was proficient at one measure within a larger economic system, they had limited to no predictive capacity at envisaging the future for the larger system.
As an example, the organisation that was best at predicting US government spending was incapable of accurately predicting what deficit the US government would run at that year — in other words, they could not predict the revenue line.
Victor Zarnowitz of Columbia University confirmed McNees and Ries’ results. He studied 70 forecasting firms between 1968 and 1982 that produced quarterly views on inflation, real growth, unemployment, nominal GNP, consumer expenditure on durable foods and changes in business inventories. His conclusion: “It is difficult for most individuals to predict consistently better than the group. For most people most of the time, the predictive record is spotty, with transitory spells of high accuracy.’’
So how does one make money in financial markets if economic forecasting is notoriously inaccurate? We are in the Peter Lynch camp. His words echo in our corridors: “Behind every stock is a company. Find out what it’s doing.... If you don’t study any companies, you have the same success buying stocks as you do in a poker game if you bet without looking at your cards.”
We have learnt from experience that the best payback on our time is to build intimate knowledge about the businesses in which we intend to invest, rather than on trying to predict transitory economic cycles. The fact is that great businesses survive all types of cycles and energy is better spent deciphering which businesses are actually “great”. That is hard enough.
But investors cannot completely ignore future economic and social forces, especially when building asset pricing models that demand economic inputs. These models are particularly relevant for fixed income teams that value bonds. Bond valuations are important, as they act as a reference point against what other asset classes such as equities are worth.
With respect to being a successful equity investor, while understanding the qualitative aspects of a company creates the foundation for long-term investing it is equally important to value the business appropriately to not overpay for it. Hence here too valuation models may include economic inputs from forecasters.
Though the work of McNees, Ries and Zarnowitz is discouraging about the forecasting ability of individual forecasters, they are sanguine on consensus forecasts. In other words, better forecasting accuracy occurs using the average of the forecasts from a group of forecasters (consensus) than any single forecast.
Echoing the studies of McNees, Ries and Zarnowitz, we have found that the efficacy of outcomes improves using two strategies. The first is inputting consensus economic forecasts into our asset pricing models. Considering that all professional investors have access to the forecasts of economists, using the average should in theory lead to asset price valuations equal to where the market is pricing these. Ironically, this is often not the case, creating obvious opportunities.
The second is an acceptance that the future is uncertain and modelling to achieve pinpoint accurate asset price valuations is impossible. A smarter approach is to create a range of valuations on an asset class or specific security, and this is achieved by incorporating the inputs from negative, neutral and bullish forecasters.
Using this approach allows us to populate our buy lists with bear, base and bull valuations for every asset in which we can deploy our clients’ capital. If a security or asset class is priced close to our bear case, it implies it is cheap. If priced close to our base case, it is fairly priced, and if near our bull case that security or asset class is expensive. This creates a highly informative framework that can be trusted to generate compounding returns over time.
A great deal of skill will be harnessed in the weeks ahead as individual forecasters prognosticate for 2025. Their work is invaluable to professional asset management firms valuing asset classes and securities on behalf of their clients.
• Clayton is chief investment officer at Northstar Asset Management.
Support our award-winning journalism. The Premium package (digital only) is R30 for the first month and thereafter you pay R129 p/m now ad-free for all subscribers.
ADRIAN CLAYTON: Best payback is to build knowledge about businesses in which to invest
Findings of studies on the predictive efficacy of forecasters are grim
Forecasters may tell you a great deal about the forecaster; they tell you nothing about the future. — Warren Buffett.
Typical for the start of a new year, a conveyor belt of leading economic and market clairvoyants (forecasters) prognosticate on everything from gin prices to stock market outcomes. Despite the endorphin release from listening to their prophecies, do they add value? Do forecasters have real predictive power, and can you make money from their advice?
Buffett is clearly a nonbeliever, but then he is not always right either. Academic literature covers the topic well. Various studies have focused on the predictive efficacy of forecasters. Unfortunately, the findings are rather grim.
Stephen McNees and John Ries, economists at the Federal Reserve of Boston, reviewed the performance of a group of leading forecasting organisations based on their quarterly forecasts of economic variables between 1971 and 1983.
This study showed that the best forecaster in one year did not repeat their success in the next. Forecasters demonstrated an inability to forecast across a range of variables, and even when a specific forecaster was proficient at one measure within a larger economic system, they had limited to no predictive capacity at envisaging the future for the larger system.
As an example, the organisation that was best at predicting US government spending was incapable of accurately predicting what deficit the US government would run at that year — in other words, they could not predict the revenue line.
Victor Zarnowitz of Columbia University confirmed McNees and Ries’ results. He studied 70 forecasting firms between 1968 and 1982 that produced quarterly views on inflation, real growth, unemployment, nominal GNP, consumer expenditure on durable foods and changes in business inventories. His conclusion: “It is difficult for most individuals to predict consistently better than the group. For most people most of the time, the predictive record is spotty, with transitory spells of high accuracy.’’
So how does one make money in financial markets if economic forecasting is notoriously inaccurate? We are in the Peter Lynch camp. His words echo in our corridors: “Behind every stock is a company. Find out what it’s doing.... If you don’t study any companies, you have the same success buying stocks as you do in a poker game if you bet without looking at your cards.”
We have learnt from experience that the best payback on our time is to build intimate knowledge about the businesses in which we intend to invest, rather than on trying to predict transitory economic cycles. The fact is that great businesses survive all types of cycles and energy is better spent deciphering which businesses are actually “great”. That is hard enough.
But investors cannot completely ignore future economic and social forces, especially when building asset pricing models that demand economic inputs. These models are particularly relevant for fixed income teams that value bonds. Bond valuations are important, as they act as a reference point against what other asset classes such as equities are worth.
With respect to being a successful equity investor, while understanding the qualitative aspects of a company creates the foundation for long-term investing it is equally important to value the business appropriately to not overpay for it. Hence here too valuation models may include economic inputs from forecasters.
Though the work of McNees, Ries and Zarnowitz is discouraging about the forecasting ability of individual forecasters, they are sanguine on consensus forecasts. In other words, better forecasting accuracy occurs using the average of the forecasts from a group of forecasters (consensus) than any single forecast.
Echoing the studies of McNees, Ries and Zarnowitz, we have found that the efficacy of outcomes improves using two strategies. The first is inputting consensus economic forecasts into our asset pricing models. Considering that all professional investors have access to the forecasts of economists, using the average should in theory lead to asset price valuations equal to where the market is pricing these. Ironically, this is often not the case, creating obvious opportunities.
The second is an acceptance that the future is uncertain and modelling to achieve pinpoint accurate asset price valuations is impossible. A smarter approach is to create a range of valuations on an asset class or specific security, and this is achieved by incorporating the inputs from negative, neutral and bullish forecasters.
Using this approach allows us to populate our buy lists with bear, base and bull valuations for every asset in which we can deploy our clients’ capital. If a security or asset class is priced close to our bear case, it implies it is cheap. If priced close to our base case, it is fairly priced, and if near our bull case that security or asset class is expensive. This creates a highly informative framework that can be trusted to generate compounding returns over time.
A great deal of skill will be harnessed in the weeks ahead as individual forecasters prognosticate for 2025. Their work is invaluable to professional asset management firms valuing asset classes and securities on behalf of their clients.
• Clayton is chief investment officer at Northstar Asset Management.
PODCAST | A closer look at the US: economic, political and market outlook
BROKERS’ NOTES: Catch a Tiger by the tail, lock in flavour at AVI
SA needs to convert the uptick in investor sentiment into tangible wins
YOUR MONEY: How to think about access bonds
SIMON BROWN: Don’t peg your hopes on value
HEATH MUCHENA: From volatility to value — reimagining portfolios with crypto
Would you like to comment on this article?
Sign up (it's quick and free) or sign in now.
Please read our Comment Policy before commenting.
Most Read
Related Articles
STOCK WATCH: Pick n Pay scrambles to turn the corner after Boxer IPO
THE FINANCE GHOST: Property sneaks the win in a close finish
SIMON BROWN: Review your stocking over Christmas
PODCAST | A closer look at the US: economic, political and market outlook
BROKERS’ NOTES: Catch a Tiger by the tail, lock in flavour at AVI
SA needs to convert the uptick in investor sentiment into tangible wins
Published by Arena Holdings and distributed with the Financial Mail on the last Thursday of every month except December and January.