Categories

Pragmatic Capitalism

Capital for Living a More Practical Life

A question on you counter cyclical portoflio allocation paper

« Back to Previous Page
0

Given the biggest draw downs occur during recessions, how does counter cyclical strategy consider recession likelihood? Or do you ignore trying to forecast a recession (using yield curve, business activity and employment trends or what have you) because of the forecasting risk from a wrong call.

That is, does this strategy suggest by allocating in a counter cyclical manner, the portfolio will be less exposed to risk on assets and thus will face lower drawdowns if a recession does occur then a 60/40 portfolio.

Closed
Marked as spam
Posted on 06/12/2016 10:53 PM
100 views
Private answer

Hi DM,

A pure countercyclical indexing strategy would ideally be systematic and involve no discretion across time as to the allocation. It would be run just like a standard indexing strategy with the exception that you aren’t rebalancing back to a fixed allocation, but instead rebalancing back to whatever the CI model says. That’s how I manage it. My opinion of the economy or markets across time have no impact on its output. It’s designed in a way that it could be 100% systematic so you take all forecasting and discretion out of the management.

A strategy involving recession prediction would entail a regime switching approach. I’ve always been intrigued by this strategy for someone who was very aggressive. For instance, let’s say you’re 25 and you know you have a super long time to leave the money in the stock market. You’d construct a 100% global equity allocation and let it ride. But let’s say you are also concerned about huge 2008 style events and worry that they could ruin your financial life. In that case you could implement a regime switch which would involve changing the allocation only temporarily when recessions occur.

Of course, the kicker is that you have to be able to predict the recession there which will sound silly to most people. It involves discretion and a predictive model which is something that isn’t proven to work well. So, it interests me, but I would be lying if I said I wasn’t skeptical of the approach. I mean, I have my own recession models, but I have no idea if they will work consistently. Still, if you think of it like insurance it might be better to have a bad model that works some of the time than no model at all which exposes you to disaster or behavioral bias….

Personally, I prefer the pure CI approach. No discretion is the higher probability outcome in my opinion.

Marked as spam
Cullen Roche Posted by Cullen Roche
Answered on 06/13/2016 12:53 AM
    Private answer

    But then how are you going to be sure that the pure CI model will work better than a standard 60/40 model.

    A few months ago you said that we are currently in the late stages of the current cycle and your model says rebalance to lower equities. Back then the S&P was at 190. Now the economy has indeed slowed but the S&P is again at all time high.

    Marked as spam
    Posted by Machine Learning
    Answered on 06/13/2016 1:02 AM
      Private answer

      The goal isn’t to “beat the market”. You don’t rebalance to beat the market. You rebalance to ensure your risk profile is aligned with your asset allocation. I could literally give zero shits where the S&P 500 is. As long as my portfolio is aligned so that I don’t get scared out of it during periods like the crash last August or the recent 20% downturn in stocks then I know I am doing something right.

      You seem to think that the purpose of CI is to generate alpha, but one of the key points of the paper I wrote is that there is no such thing as alpha in the aggregate. It is a fool’s errand to chase alpha!

      Marked as spam
      Cullen Roche Posted by Cullen Roche
      Answered on 06/13/2016 1:12 AM
        Private answer

        Thanks for the response.

        What do you make of using a simple rule like the cross of the 10 month MA as an exit from equities to de-risk downside? I have read papers (Faber) that suggest using this to complement one’s portfolio allocation rules (though it seems a sideways market like the late 60s to 80s) would have you trading in/out without much benefit. Does such a rule to limit risk add unneeded “complexity” to a CI strategy?

        Also would you consider a home be an “investment” rather then savings by your definition as one “consumes” production (i.e. lives in their home)? That said, homes like stocks (that track profits) should not be expected to provide alpha in long term as homes should track income (as well as interest rates and supply/demand) though pricing could be more “sticky” then stocks.

        Marked as spam
        Answered on 06/13/2016 2:05 PM
          Private answer

          Yeah, the problem with those momentum sort of strategies is that they tend to give you lots of false readings in real-time. They’re very hard to implement. I believe Meb implemented this strategy with his original ETF and it didn’t fair well. It’s not that the methodology is necessarily flawed. It’s just that translating this into real-life is not very easy….

          Marked as spam
          Cullen Roche Posted by Cullen Roche
          Answered on 06/14/2016 1:37 PM