Distinguishing Alpha from Noise

I read a paper today that I thought was pretty interesting — A Consultant’s Perspective on Distinguishing Alpha from Noise. [8 pages PDF]  I have been on both sides of the table in my life.  I have hired managers, and I have tried to sell my equity management services.

In general, managers that thought would offer value would venture off the beaten path.  They might own some well-known names, but they would own far more that would make me say, “Who is that?”  The companies would be less known because they are smaller, foreign, have a control investor, etc.

Those portfolios would look a lot different than an index fund.  They would be more concentrated by sector, industry and company.  They would have a process that analyzes what the market is misvaluing, whether by sector, industry, or company.  They would stick to their discipline through thick and thin, realizing that all anomalies in the market go in and out of favor.

The process would specify what anomalies of the market, or what information advantages the fund would attempt to exploit.  But once you specify that, you stick to that as your strategy.  There is no room for tossing an asset in “because it looks good.”

There is a balance in good strategies that allows for minor modifications around core principles.  All good strategies have to adapt, but there has to be a strategic core from which the strategy will never vary.  Absent that core, the strategy will give in to fear and greed — buying high and selling low.

Quoting from the paper:

I am amazed at all the managers that make an assertion of the type “In the long run X always wins”, where X could be dividend yield, earnings growth, quality of management, a quantitative factor or mix of factors, etc., yet are unable cite a reason why X should be systematically under-priced by the market.

My view is twofold.  There are some ugly situations involving financial stress that most investors don’t want to take on.  There are also less glamorous companies that few want to buy.  Those can be excellent investments.  My second point is tougher to make, but industries go in and out of favor.  So do market factors.  Buy that which is safe, and out-of-favor.

Now, for managers, I would recommend keeping a trading journal, where you record why you think your investment hypothesis will succeed.  If your investments succeed for reasons that you specified in advance, that is an indication of skill.  There is a lot of what is called “luck” in investing.  If you are beating the market, and it is not for reasons that you specified in advance, you do not have skill, you have luck, and luck strongly tends to mean revert.

My view comes down to this: I like to see a long track record of outperformance, an unusual portfolio, and a strategy that convinces me that you have discipline, and a constructive way of finding undervalued assets.   Absent that, I will probably think that you are a pretender than an outperformer.  There are always some that outperform for a short time, and then underperform as the underlying economics shift.  Markets are volatile enough that there are always some with three-year track records that are stunning, and very lucky.

Separating luck from skill — that is the toughest aspect of investing.  But it is needed because there are so many investment managers touting skill, and what do they really offer?

2 Comments

  • TMoney says:

    It occurs to me, that it is easier to separate good professional gamblers from the also-rans than it is money managers. Over the long haul investing has a positive expectation of return, unlike most gambling. Therefore, good professional gamblers MUST generate alpha or go bust. I believe successful professional gamblers would be good money managers, if they applied their talents to looking for investment opportunities in the markets instead of the casinos.
    Perhaps it is easier to find people who have the skills to manage money and hire them rather than look for people who are managing money and have skill.

  • lvl says:

    The trading journal is very important as you point out, a way to mark to market your assumptions and theses. Completely agree. However absent that, its hard to assess claims of discipline – are you disciplined because you “hold-on” by “ignoring volatility” or looking through non-fundamental dislocations, or are you disciplined because you manage risk by cutting losses often. Maybe it’s a combination and then again you need a long track record of out-performance that you predicted in advance. Otherwise you might be Phil Falcone or John Paulson – they predicted in advance and scored big, but it’s not a long track record. Which brings us to the unusual portfolio part. I’m not sure that should be a firm rule. Value minded portfolios seem fairly conventional at this point, low costs( low capital costs from say insurance premiums, or low expenses from product selection and turnover), low valuations (cashflow or EV multiples), inverse matching of operating leverage to business cycles.

    What separates various value portfolios is the discipline: define the objective, identify economic/cyclical circumstances, pose a target you wish to find, and screen for it… then track your results… eg. Momentum works too, but it’s an entirely different game and I don’t want to see it in a value portfolio. I’d rather see volatility and jumps, value should be illiquid in some sense. Otherwise my rule is that 95% of Alpha is leveraged Beta :)

9 Trackbacks