Smoothing the market for alpha

How to get money to alpha, and vice versa.

The problem

Let’s focus on two groups:

  1. People who have money and want alpha
  2. People who have alpha and want money

When there is a cross between those two groups, there is happiness.

Identifying Group 1 is easy.  Identifying people who want money is not difficult either.  Identifying people who have alpha is a bit more problematic.

A person who believes themselves to be in Group 2 needs to convince at least one person in Group 1 that they are indeed in Group 2.  How to do that?

Tool number one is a backtest.  There are run-of-the-mill backtests and more informative backtests.  But they all suffer from the possibility of data snooping.  It is hard work to assess the quality of a backtest.  Hard work is not necessarily something that members of Group 1 are keen on — and anyway they probably have plenty of work already.

Tool number two is a track record.  There are two types of track record: paper trading and live trading.  Paper trading is often regarded as a backtest by another name (sometimes correctly so regarded).

A live-trading track record can be funded by personal money in some cases.  But many of the more interesting strategies require substantial amounts to be viable.  You need to get money in order to get money.  Catch-22.

The current system of crossing Groups 1 and 2 is quite chaotic.  It depends highly on salesmanship.

Hypothesis: Salesmanship is negatively correlated with alpha.

If this market were made even a little more efficient, that would add a serious amount of value.

An idea for Bloomberg

My friend Giles Heywood (who puts himself in Group 2) has a solution. It involves Bloomberg, and would seemingly be in their best interest. (Three months ago) Giles proposed:

In PORT, the now-greatly-expanded portfolio analytics hosted on Bloomberg, they implement a simple option: that portfolios cannot be revised for past dates, and positions are implemented at VWAP in the future (rest of day, next 24 hrs, etc). This then becomes an audited track record – it excludes market impact, but otherwise is highly realistic. This portfolio’s performance and risk can then be permissioned for sharing with other Bloomberg users, who would include incubators.

From the incubator perspective, they can trawl the universe for attractive risk/return performance in the investment universe of interest, monitor positions, discuss the logic behind them, see the process in action. They reduce risks, costs, and get easy access to a large pool of talent, arguably the entire pool of interest.

From the wannabee perspective, it gets them over the chicken/egg problem: no capital > no track record > no interest > no capital (repeat). For someone like myself who sits in a HF office administering my process as a paper portfolio on an accounting system (and PORT) and still awaiting capital allocation, sitting somewhat frustrated on a deep capacity Sharpe Ratio 4 process, it could open more doors a bit quicker.

From Bloomberg’s perspective, it gets them into a very interesting space, matching supply and demand in the market for talent. Intriguingly the idea was not new to them, but mysteriously somehow had never got out of the blocks. Watch this space? Someone will do this, and it will be fascinating when they do. Lets be clear: there are already ‘fantasy portfolio’ competitions such as CNBC’s, but these are shoot-the-lights-out competitions for punters, they are not risk-adjusted, they are not testing the right metrics, they do not offer drilldown, analysis, monitoring, they are candy-floss.

Giles reports having seen significant interest from people at Bloomberg, but no action so far.  It seems like a bit of lobbying could produce action.

A concern that has been expressed is the possibility that people would put up a large number of strategies and depend on one of them randomly doing well.  I wouldn’t think that should be a problem if others are able to see how many strategies an individual has on the system.  Anyone dodgy enough to circumvent that check is probably dodgy enough not to pass other due-diligence checks.

Still missing

To really assess the performance of the strategies, members of Group 1 should compare the strategy to random portfolios that have the same constraints.  That seems too complicated to be incorporated into Bloomberg (though I’d love to be proved wrong).  But strategy owners could run random portfolios themselves to show to potential suitors.


I’m looking at this as an outsider (I’m in neither Group 1 nor Group 2).  What have I missed?


“This stuff is better than cotton candy, really it is. It’s made out of real cotton. Yossarian, you’ve got to help me make the men eat it. Egyptian cotton is the finest cotton in the world.”

from Catch-22 by Joseph Heller

Subscribe to the Portfolio Probe blog by Email

This entry was posted in Fund management in general and tagged , . Bookmark the permalink.

Leave a Reply

Your email address will not be published.