There has been some chatter in the economics universe in recent days about the usefulness of economic forecasting (see here and here for instance). The concept of “forecasting” is highly controversial in both economics and finance. Some say we shouldn’t do it. Some say it’s just impossible. Some say it’s possible. Some say it’s just an imprecise part of life as forward looking animals. I fall into the last category there. Forecasting isn’t just a reality as forward looking animals, but it’s a crucial element of our existence and our very ability to survive.
I like how Simon Wren-Lewis refers to forecasting as “intelligent guesswork” as opposed to “forecasting”. Forecasting implies crystal ball gazing or something like that when the reality is that we should just be trying to understand the potential impacts of certain economic and market events so we can create a high probability of creating successful policies or positioning ourselves so we’re prepared for the potential risks and rewards that might arise. It’s sort of like preparing to set sail across an ocean. You would never claim to be able to predict the weather or conditions you might encounter, but you can understand your path, potential weather patterns and how your vessel operates so you can increase the odds that you will make if there in one piece.
Economists and market forecasters shouldn’t bother with “forecasting” or fortune telling. They should construct models of the world that simply increase the odds of understanding and applying potential future outcomes. I call it probabilistic forecasting. You don’t need to predict the future. You don’t need a crystal ball. In fact, probabilistic forecasting assumes you will be wrong often. It assumes the system is dynamic and complex. But it doesn’t assume it is totally random or totally unpredictable.
There are several problems that play into some people’s aversion against forecasting of any type. The most glaring is the poor track record of economists in predicting, well, most things. Let’s face it – economists of all stripes have a pretty piss poor track record predicting the future. Of course, if you’re a regular reader of this site then you probably know that there’s a lot in mainstream economics that I just don’t agree with. So it doesn’t surprise me that bad models and bad understandings have led to bad predictions. The most glaring example of this in recent years was the use of the money multiplier when accounting for potential QE outcomes. Dozens of famous economists and market pundits said QE posed some huge inflation risk as the reserves would be “lent out” at some point in the future (see here for instance). Of course, their understanding of banking was flawed so the underlying understandings led to bad predictions.
This is a great example of how probabilistic forecasting can be utilized assuming one has a sound understanding of the monetary system. No one using an endogenous money framework for understanding the monetary system would have predicted high inflation, high interest rates or anything like that after QE because it didn’t mesh with the underlying understandings. Those of us who got this right didn’t make great predictions there. We made what looked like highly probable forecasts about the direction of interest rates and inflation based on a sound understanding of the monetary system.
One could argue that forecasting is even more important for policy as it effects so many people at the macro level. For instance, countless politicians have argued in recent years that the high levels of government spending and debt posed a serious risk of the USA becoming the next Greece. These forecasts were repeatedly used to justify spending cuts and tax increases that almost certainly hurt the recovery. But these forecasts were again based on misunderstandings. Anyone who had a sound understanding of the US monetary system’s design knew that Greece was fundamentally different and inflation never posed a serious threat given the unique de-leveraging environment we were in. But again, poor understandings of the monetary system led to bad forecasts which hurt millions of people. These low probability predictions dominated the policy discourse and led to bad decisions based on unwarranted fears.
Some people on the market side of things will argue that we shouldn’t bother making market forecasts either. But this is a naive view of the world. As I’ve explained before, anyone making a directional bet on certain asset classes is making an implicit underlying macroeconomic forecast. The fact that some investors choose not to change their forecasts over time doesn’t mean they are engaging in “forecast free” investing. It just means that their forecast is static. For instance, a long only stock focused indexing approach is based on an implicit bullish economic forecast. If the underlying economy doesn’t perform well over the chosen investment period it’s highly likely that the underlying stocks won’t perform well either. All you’ve done is make a bullish macroeconomic bet through stock prices. In my opinion, you should be fully aware of the underlying forecast that certain asset classes are dependent on for their expected success in order to construct a sound portfolio. So again, you see the importance of forecasting when applying our understandings to actions.
All in all, forecasting isn’t just a reality of economics and finance. It’s a crucial piece that we all rely on daily whether our forecasts change or not. Some of those forecasts are good and some are bad. Some are based on political ignorance while others are based on rational probabilistic understandings. But we need those forecasts in order to establish even a general sense of the future. It’s okay that we make forecasts. But we should demand that those forecasts be less dogmatic than they generally are and we should demand that they take root in empirics rather than politics. No one’s a fortune teller. No one knows the future precisely. But that doesn’t mean we have to go through life in a dogmatic haze pretending it’s all just a big random walk that is completely unpredictable.