S&P with 10 and 40 week overlays

Since 2000, this has given profitable signals 6 of 7 times

So a year ago, I was thinking about how to best leverage all the academic papers that come out showing their research that says x is better than y and that if you apply this or that you will get better returns.  Unfortunately, so much of the research is not particularly useful in the real world.  So culling through the papers, finding trends and general ideas that do seem to hold up is a bit of a task.

A couple of ideas that seem to be “tried and true” at this point are:

  1. the benefits of diversification
  2. the existence & advantage of momentum
  3. returns are inversely correlated with trading activity

I’ve tried to keep each of these in mind when creating my tactical asset allocation system.

The well diversified portfolio, holding uncorrelated instruments where possible limits portfolio draw-down while increasing returns.  However, too much diversification, and you are just holding the market portfolio.  So there is a balance.  My personal balance is to weight some sectors more than others, and not hold some at all, depending on performance and other variables.

Returns for TAA over S&P

Strategy vs. S&P 500 returns

Many papers note the existence of momentum, and that there are lasting and outsized benefits derived from prior performance.  The relative strength rotations model I developed, again inspired by Mebane Faber of The Ivy Portfolio fame, more heavily weights the holding of better performing assets compared to lessor.  The old adage of cut the losers and let the winners run is a good guide to follow.  However, this momentum idea can get very complex very quickly, so limiting it to simple price momentum over a couple of different periods is what I focus on.  No double or triple filters, no cherry picking periods, etc.  Keep it simple and (relatively) easy.

The return chart to the left is a comparison to the simple 10/40 week moving average crossover compared to the S&P 500 since 2000.  Yes this is cherry picking the timeframe, but the best I could do right now.  You own the S&P when the 10 week moving average is above the 40 week, and are in cash if the 10 week is below the 40 week moving average.  Very simple system, and shows that being in the market all the time can be dangerous.

Equity Curve for 10/40 moving average crossover

Equity Curve for 10/40 moving average crossover

On a related note, I have a relative strength rotational model for investments that I have created.  The image is an example of this type of system.  If you’re interested, read more about my tactical asset allocation investment model based on relative strength, good diversification, and market timing stops.

Tactical asset allocation portfolio - relative strength ranking

Relative Strength rankings

The last item that I saw in research and other brokerage based surveys, are that returns are inversely correlated with trading activities.  In other words, more trading hurts performance.  I think all the various brokerage ads where they give you tools to trade quicker and quicker from mountain tops and while driving from your mobile phone all point to there being a conflict of interest.  They make money on your trades whether you do or not.  Once again, balancing this with cutting losers and letting winners run is key.

The last point I want to touch on is all those stock advisers saying if you missed the best 10 days of the market you would only earn half the returns, and since you don’t know when these will occur you always have to be in the market.  What they don’t tell you is what would happen if you miss the 10 worst days, or both the 10 best and worst.  The last link below explores some of this sales talk and gives you the real story.  This is the basis for the timing aspect to my TAA model, which keeps me out of the large declines.

Research Papers:

Mebane Faber, Ivy Portfolio basis

Mebane Faber Relative Strength

Mebane Faber, Where declines hide

Trading & Costs

Readers, what do you think about using the latest research to try and juice your portfolio?  What about the

About Karl

has written 56 articles on this site..

11 Responses to Academic Investing Advice – What do PhDs know that you don’t?

  1. I think constantly going in and out of market trying to time the market is a losing proposition. Just buy solid stocks and hold on to them for many many years. It is difficult to actually make money in the stock market and many people try it thinking they could make large profit.

    • Karl says:

      I agree with you under certain conditions. The primary condition being that we are in a multi-decade bull market. In that case, holding any and all stocks is a winning position and any other strategy that gets you out of the market will underperform. However, if you are in a bear market or a wide trading range market (1960-70’s and 2000’s) then being out of the market at times dramatically increases returns. I added a return comparison and a note above addressing this.

  2. bax says:

    Good points you raised there, I always wind up fighting my stupid emotions of fear and greed when I try to time anything to do with the market.

  3. YFS says:

    Many analyst focus on the fact of missing the best 10 days because no one has consistently timed the market. It’s easier to buy/hold to ensure you don’t miss the best 10 days. How has your TAA portfolio done in the last 5 years compared to the S&P 500?

    • Karl says:

      YFS, the research done by Faber indicates that between 66% and 75% of the best return days occur when the market is under the 200 day MA. Given that usually denotes a bear market ahead of time, its fairly easy to avoid. Backtested results for 5 years are very good, beating the market by some 30%. Recent years have seen slight under-performance because of the strong bull.

  4. The momentum theory makes sense, but it got complicated pretty quickly. I need to learn more about moving average. For now, I’m more of a buy and hold investor.

  5. WealthNerd says:

    I very much enjoyed the level of detail and sophistication in this article! Being a very analytical investor myself, I can appreciate all of the data that you’ve presented! (I’m currently involved in academic research pertaining to machine-learning for autonomous equity trading and portfolio optimization, and on my own time I develop models for hedging risk using options and statistical arbitrage)

    I will say this: you won’t find very much “academic” information about well-performing strategies/techniques (at least anything that applies to the present), as it’s largely a very “hush hush”, proprietary industry. However, that’s not to say that such strategies don’t exist. Most investment banks and hedge funds make on the order of hundreds of millions per year by such methods.

    Active trading doesn’t *necessarily* hurt your long-term performance, but it does tend to be correlated with stock market psychology in a destructive way (ex: people feel that they “need to be trading”, and the impulsiveness makes forced bad decisions). If you can remove emotion from the equation, your odds are far better. Particularly, a business partner of mine and I created a grey-box algorithmic trading system that implemented a trend-following strategy using real-time tick data (on the order of seconds) and traded on 5-minute bar intervals. Based on a number of parameters that involve purely momentum (zero fundamental analysis), it would prioritize “good plays” based on the attractiveness at the time. The time between entering a position and exiting a position tended to be somewhere between intra-day and intra-week; in a nutshell, it made the picks for us, although ultimately we made the final decision ourselves. I will say that its historical performance (as well as realized performance) was pretty good, mostly irregardless of market conditions (can make money on the way up and on the way down).

    I’ll leave it at: you can make some pretty stellar returns with minimal risk despite your trading frequency if *you’re properly hedged*! Chasing fast movers on its own isn’t much of an insurance policy on the long-term, as we’ve seen by the law of large numbers. 😀

    Keep up the reads!

    • Karl says:

      WealthNerd, I appreciate the detailed response. Definitely the correlation to trading activity and losses is for the general investing public, not the HFT that is specifically designed for that. I’ll also agree that the real money-making strategies won’t be released in any literature or academic papers. Far more to be made in selling to a fund or bank or starting your own thing.

  6. […] you like detailed analysis? CoW newcomer Karl at Cult of Money has some for you. It’s somewhat technical, and a glimpse into the mind of someone whose […]

  7. Stan says:

    any way to subscribe to your posts by email?

Leave a Reply

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

shares