Tuesday , March 26 2019
Home / Cullen Roche: Pragmatic Capitalism / The Counterproductive Nature of Annual Forecasts

The Counterproductive Nature of Annual Forecasts

Summary:
The annual parade of Wall Street forecasts are starting to rollout. I will be honest – I love these things like a fat kid loves cake. It’s fun to try to predict and plan for the new year and there’s no better way to do that than to read what a whole bunch of smart people are predicting. The problem is, annual forecasts are worthless and I would argue even counterproductive. Let me explain why. I’ve talked a lot about how I like to think about all financial assets as bonds. I know, I know, stocks aren’t bonds, but the math in bonds is super clean and it makes things simple to compartmentalize. For instance, if I buy a AAA rated 10 year T-Bond I know many useful things: Its time horizon – 10 years. Its annual income – 2.9% as of this instant. Its duration – 8.62. This is very useful

Topics:
Cullen Roche considers the following as important:

This could be interesting, too:

Cullen Roche writes An Open Letter to the US Sentate – Vote No on Stephen Moore

Cullen Roche writes The Shortest (and Best) MMT Primer You’ll Ever Read

Cullen Roche writes Is Government Debt “Equity”?

Cullen Roche writes Reconciling Krugman vs Kelton

The annual parade of Wall Street forecasts are starting to rollout. I will be honest – I love these things like a fat kid loves cake. It’s fun to try to predict and plan for the new year and there’s no better way to do that than to read what a whole bunch of smart people are predicting. The problem is, annual forecasts are worthless and I would argue even counterproductive. Let me explain why.

I’ve talked a lot about how I like to think about all financial assets as bonds. I know, I know, stocks aren’t bonds, but the math in bonds is super clean and it makes things simple to compartmentalize. For instance, if I buy a AAA rated 10 year T-Bond I know many useful things:

  • Its time horizon – 10 years.
  • Its annual income – 2.9% as of this instant.
  • Its duration – 8.62.

This is very useful information. I know my exact return over an exact period of time and I can calculate the asset’s sensitivity to future price changes. With an instrument with near zero default risk you don’t need to know much else to make smart decisions.

The last figure above is really, really handy because it puts the risk of the instrument in perspective. I like to think of duration like the leash on my dog when we go for walks. The farther the dog is out on that leash the more risk there is that she’s gonna do something I can’t control. The more I tighten it up the less exposed I am to her doing something I can’t control. But I’ve gotta have some duration otherwise we’re not walking – I am just carrying a fat dog around like an idiot. Same basic thing applies to financial assets – you need some duration, but you need the right amount of duration.

Now, stocks are a bit different from bonds, but you can apply the same basic concepts. For instance, I’ve calculated the duration of the stock market at 28. This means that, at current yields, it would take about 28 years for the stock market to recover from a 50% decline. That sounds like a lot. Maybe it is. William Bernstein and John Hussman calculate it at even longer. There’s obviously a lot of guesswork going on here, but the point is that the stock market, much like the bond market, is reasonably viewed as a fairly long-term instrument because it takes time for its underlying entities to pay out the cash flows that we see as consistent future returns.

Let’s be somewhat conservative and assume that the stock market will generate its 7% annual returns paid out entirely as dividends and that valuations will compress a bit bringing our average annual returns to 5%.¹ In this case our duration is 10 years. That seems pretty reasonable to me since it’s unusual to see any 10 year period in which stocks are losers. So, the stock market is at least a 10 year instrument and the bond market, in aggregate, with an average duration of ~7 gives us a 50/50 stock/bond portfolio with an average duration of 8.5. So, why in the world would we judge portfolios with 8.5 year time horizons over one year periods?  The answer is that we shouldn’t, but that we do it because we’re really f&$king impatient animals. 

This makes sense from a forecasting perspective as well. We know, statistically, that the stock and bond markets have positively skewed long-term returns. The more patient you are, the more likely you are that an optimistic forecast will be realized. The more short-term you are, the more random the outcomes and the less likely your forecasts are to come true.² And that’s the real problem with annual forecasts. The stock and bond market returns are largely random inside of any 12 month period. So it makes little sense to try to guess what the stock and bond markets will do inside of any single year.

So, instead of thinking in one year time horizons try to treat the financial markets for what they are – moderately long-term instruments where reasonable forecasts can only be relied upon over multi-year periods. This will help you avoid the bias of short-termism and increase the likelihood that you’ll achieve your financial goals by avoiding the urge to constantly change what you think isn’t working because you’re too impatient to let it work.

¹ – I assumed a constant profit rate consistent with historical rates of profits. I also assumed a bit of valuation compression since valuations are a bit high in historical terms. 

² – Investing is all about duration matching. You need to quantify your personal duration for your financial goals and match the right portfolio to that duration. 

Cullen Roche
Former mail delivery boy turned multi-asset investment manager, author, Ironman & chicken farmer. Probably should have stayed with mail delivery....

Leave a Reply

Your email address will not be published. Required fields are marked *