A powerful two step process to deal with the endowment effect: The game of two thirds, or how to deal with free loaders in your portfolio

Would You allow tenants to stay rent-free ?

Would You allow tenants to stay rent-free ?

If You were the owner of an apartment building, would You allow tenants to stay rent-free forever ? You would probably do everything in your power to either collect or evict free loaders. In the investment realm however, one of the main reasons managers fail to accomplish their goals is that they allow free-loaders to stay rent-free in their portfolios. The difficulty then is how to identify and deal with free loaders.

They don’t really stand out enough on an individual basis. Yet, as an aggregate free-loaders put a drag on performance.

  • Endowment effect: once in the portfolio positions are sticky.
  • How to identify and effectively deal with free loaders
  • The 3 main benefits of the game of two thirds
Once upon a time, i used to place thematic small positions across the portfolio like pawns on a chessboard. They were supposedly hedges for China, precious metals, oil, monetary intervention etc. They were all tiny positions that were supposed to kick in if any of these themes were to gain traction. Six months went by and I could not understand why performance was so pedestrian. Meanwhile, none of those stocks had worked. Then, it dawned upon me that even though they were tiny individual positions, they totaled 10% of the portfolio as an aggregate.
The endowment effect (Thaler 1980)
Endowment effect is the hypothesis that people value more what they own than what they could buy. It is hard for positions to dribble their way into our portfolios, but once in they become sticky. It is difficult to get rid of them, even though they do not contribute. Some managers would hold on to losers just because they do not know what to buy next.
Our lives, our desks, our houses are filled with clutter. Unless we actively create and enforce rules to get rid of it, clutter creeps up on us. Our inner saboteur will always find good reasons to hoard junk. To illustrate its potency, let’s look at a simple example: in your wardrobe, isolate the clothes You have not worn for over a year. Think about all the excuses to keep them, but then ask yourself: “If i did not have it, would i buy it now ?” If not, then bye bye, fashion moves on and so should You.
The game of two thirds: A simple two-step process to deal with free loaders
Free loaders neither detract nor contribute enough to be visible. They don’t stand out enough to be dealt with. Since it is not possible to deal with them on the y-axis (price), the solution is to introduce time x-axis. Rationale is simple, if stocks have been there for some time, but still fail to contribute, then their weight should be reduced.
  1. Calculate portfolio turnover, divide it by three: first 1/3. Add 1/3 turnover to the entry date of each position. For example, a stock entered on January 5th and a turnover of 1 would yield a cut-off date of April 5th
  2. Divide performance in 4 quartiles, concentrate on the third quartile: second 1/3. For all stocks in the third quartile past their anniversary date, cut weight in half
 Special mention for long-term winners
Apple (AAPL) or Softbank (9984:JT) are long-term winners. They sometimes go through extended periods of under-performance. Because there is so much embedded profit, it is difficult to realise that they have not contributed for some time. The idea then is to reset contribution on a rolling basis.
The idea then is to apply the same rules as above on a rolling basis. Instead of cutting positions to half, taking a portion the size of the out-performance from the previous haircut. For example, if Apple went up by 10% from previous haircut, then shave 10% off the current size.
The rationale is
  1. if it starts to underperform, it will be dealt with, and this profit taking will have cushioned the blow. This demonstrates stewardship
  2. If it continues to go nowehere, resources are re-allocated to a potentially more productive asset. If non-performance persists over 2/3 of portfolio turnover, then a more drastic reduction is in order
  3. if outperformance resumes, then it will be dealt with
It is important to periodically reset contribution. When stocks have been in the portfolio for a long time and substantially contributed, we become attached. Failure to reset contribution is one of the reasons why some managers escort their positions on the way up and then all the way back down. It doesn’t show until it is too late.
The three benefits of the game of two thirds

The game of two thirds may appear simplistic. It has however powerful psychological implications. It is a simple, powerful and objective way to short-circuit the endowment effect for three reasons:

  1. Simplicity: math is beyond dispute. Simple rules are elegant, easier to implement and harder to challenge
  2. Stewardship: great investors are not smarter, they have smarter trading habits. Getting rid free loaders builds the habit of dealing with difficult stocks
  3. The quality of our excuses determines the quality of our performance: one of the most frequent excuses is “what do i buy next ?” Constant re-examination of positions forces managers into action.
Once in a portfolio, positions are often sticky. Asking ourselves “would you buy it again today ?” is too subjective to deal with positions that have overstayed their welcome. Our inner saboteur will find good reasons to procrastinate until the next review. Our natural instinct to hoard junk “endowment effect”. The game of two thirds is an elegant way to identify and deal with free loaders.


4 replies
  1. Derek says:

    Hi, I am wondering how you calculate the turnover by time. Usually I think about turn over as some $ amount bought or sold / NAV of the portfolio. Can you give a more detailed example of calculating the turnover of a multi asset portfolio with different start dates for positions? I would like to see how to put this more into practice thanks.

    • lbernut says:

      Hello derek,

      Happy New Year.

      Thank You very much for your question. I assume your question refers to the treatment of options and leveraged instruments in a portfolio that also contains cash securities. Basically, how do You treat leverage, moneyness and time-value ? Is that correct ? Thorny issue indeed.

      In terms of cash outlay, OTM options represent a tiny fraction of turnover, but they command massive underlying. We have often used delta-hedged 3M notional exposure for slightly OTM options. The difficulty is time-adjustment.

      If the question is related to different start date positions, then the question has more to do with the calculation of AUM. In this case, modified-Dietz is the de-facto time series standard for AUM calculation.

      Personally, I pay more attention to net trading than turnover Net Short trading = ($ Mkt Val Buy-To-Cover[n] – S Mkt Val Sell Short[n]) /AUM[n] and similarly Net Long trading = ($ Mkt Val Buy[n] – S Mkt Val Sell[n]) /AUM[n].
      I plot this as column in a chart that also contains net exposure, gross exposure and market. This is my “who are You ?”. Market went up or down, what di d You buy and sell which resulted in what bullish/bearish view (net exposure) and how much confidence (gross exposure)

      Hope it helped at least a little bit

      Meanwhile, Happy New Year

  2. ks says:

    Hi, was just wondering why are you using the second 1/3 in the third quartile ? And if any stock passed its anniversary date wouldnt you anyways cut it off??

    • lbernut says:

      Good morning Karan,

      3rd quartile performers

      This game is designed to help discretionary investors. I am 100% systematic, 100% algorithmic and automated. I have worked with discretionary market participants most of my career. I therefore empathise with their challenges. Since i love formalising stuff, i came up with a simple rule to deal with time wasters. This is not meant to be optimal or even optimised. Concept is there, it makes sense and it is easy enough to enforce. That is all that matters
      Discretionary investors usually appreciate simple user friendly rules. The mnemonic here are simple: isolate the 3rd quartile performers. On the 1/3 anniversary date, cut position size in 1/2.

      Secondly, time wasters are good hiding. So, the whole purpose of the exercise is to uncover them

      Why 3rd quartile then?
      1st and 4th quartile performers are highly visible: they seat at both extremes of the performance distribution
      2nd quartile are solid but less visible contributors
      3rd quartile are eclipsed by 4th quartile. One of the blindspots we have the contrast effect between really toxic stocks and ordinary bad stuff. The best way to identify time wasters is to segment their population, hence the use of quartiles

      Why 1/3rd of turnover then? Time is money, it compounds.
      Every time we say Yes to a loser, we say No to a potential winner. So, the idea is to give to stocks enough time to mature 1/3 of turnover, but keep it short enough for recovery (2/3 of 1 turnover left) if it did not work out.
      This can be done over different time periods: 20-30-50-100. I would be weary of optimising this though. Market regimes change over time and so does our response

      Why halve the position then?
      Real estate is always proportional to rent. What doesn’t contribute should not occupy much real estate.
      In the end, the smaller the position, the easier it is to close. There is a size effect here.

      What to do with the freed-up resources?
      The best thing is to allocate to a new idea. For some professionals, turnover is perceived as a bad thing, simply because it is a recurring question. The great shoe designer Michael Jordan said that he missed 100% of the shots he did not take.
      The second best thing is to allocate to 1st and 2rd quartile positions: ride winners, cut losers

      I hope it helped. Let me know what You think


Leave a Reply

Want to join the discussion?
Feel free to contribute!

Leave a Reply

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