Thread by @ritholtz

Professional Money managers know how to buy. They are actually pretty good at it. What they need to do is to learn how to sell. They are mostly terrible at it. By @alexoimas +3 others 1/

bloomberg.com/opinion/articl…

Doubt that? This clever, insightful research paper is required reading for anyone who manages money professionally or has their own cash invested.

It hits all the right notes: good data, robust methodology, clever analysis, big, bold conclusion. 2/

papers.ssrn.com/sol3/papers.cf…

The shocking punchline: Fund managers would be better off, and in some cases much better off, SELLING HOLDINGS AT RANDOM rather than choosing them the way they usually do.

Random selling beat managers sells by 50-100 bps gain over the following year 3/

That is astonishing.

Money managers exhibit skills + generate alpha in stock picking abilities. But they give it up + more in their sell decisions. Even relative to strategies involving no skill at all like randomly selling existing positions 4/

ritholtz.com/2019/01/bbrg-s…

Why? The behavioral Asymmetry between buying and selling.

Purchase decisions are forward-looking;
Selling decisions are backwards-looking - these are more susceptible to behavioral biases / cognitive errors we typically think of as common among non-professional investors. 5/

I guess I should not be surprised, but: the study found professionals were just as likely to suffer from biases, heuristics and behavioral errors as amateurs.

The pros are just as bad at managing themselves as mom & pop investors 6/

The study looked at 783 portfolios, average value of $573 million, from 2000 to 2016, with more than 89 million trading data points, 4.4 million trades (2.0 million sells, 2.4 million buys of veteran institutional portfolio managers. 7/

Assets with extreme returns are more than 50% more likely to be sold than those that just under- or over-performed.

PMs sold stocks that had risen a lot, ignoring Momentum as a return factor
They sold stocks that had fallen a lot, ignoring Value as a return factor

8/

The researchers created a "counterfactual portfolio with random sells.

It outperformed portfolios with PM determined sells by 50 to 100 basis points over the following year.

Random selection did a better job of keeping winners and tossing losers than the fund manager did. 9/

When the researchers created a similar counterfactual Buy portfolio, it underperformed the active strategy.

Thus, PMs demonstrated genuine skills — but only when making their buys; little or no skill when making their sells. 10/

Consider the profound ramifications this has for active managers.
It shows a path towards improvement, to regain market share + fee-generating assets vs the trend toward passive indexing.

Better become much more skilled and disciplined at selling, and soon! 11/

Ironically, traditionally negative academic criticism about active management ( according to findings of this research) might actually UNDERSTATE the extent of the active manager’s performance problem.

The 4 key premises to traditional academic theories . . . 12/

• Markets are efficient zero-sum games with winners + losers

• Biases + cognitive errors impact manager decision-making

• Winners' advantages fades to zero as drag of fees + trading costs compound

• Identifying out-performing managers in advance is mostly impossible 13/

We can now add one more element to the above: managers lack of skill when selecting what to sell of their holdings.

The most common reason given to sell was to buy the next “great” idea.

This has been a disaster for both returns & active management business models. 14/

Biases are everywhere:

Purchase decisions are forward looking while sales are backward looking.

It is the backwards looking process that seems to be most ripe for the behavioral biases and cognitive errors exhibited by professional and non-professional investors alike. 15/

4 final observations:

1. Sophisticated professional market participants suffer biases + employ heuristics similar to those of retail investors / traders.

2. This reflects an asymmetric allocation of cognitive resources, particularly attention, when it comes to selling. 16/

3. Managers have substantially greater propensities to sell positions with extreme returns:

Both the worst and best performing assets in the portfolio are sold at rates more than 50% percent higher than assets that just under or over performed; 17/

4. PMs do not lack the fundamental skills to sell well; when relevant information is readily available – e.g., earnings announcement days – they can effectively incorporate data into their selling decisions. This outperforms the randomized selling 18/

Now for the bad news: Behaviorists like Danny Kahneman and @R_Thaler have observed how difficult it is to manage your own biases. The Nudge idea is to deploy systems that operate with your biases leading to facilitate better decision-making. 19/

Unless and until Active managers significantly improve their selling discipline, they are likely to continue to lose market share to passive indexing. /END