Inevitable Automation Of The Fund Management Industry Is Upon Us

| About: Financial Select (XLF)
This article is now exclusive for PRO subscribers.

Originally posted on TheStreet on January 4, 2017

Why do human fund managers still exist?

This might seem like an odd question, but in light of a range of critical factors, some old but many new, there is very little economic rationale for these managers to exist, except perhaps in rare cases.

For starters, there is the poor or at best average historical performance of most discretionary fund managers, including alternative managers such as hedge funds and private-equity houses, certainly on a risk-adjusted basis. Much economic research has continually proven the rather harsh adage that having someone throw darts at a list of S&P 500 stocks produces the same result as that of the average fund manager.

To put it another way, very few fund managers outperform the market in the medium to long term. But given the difficulty of trying to predict the future, this isn't really something that any mortal fund manager could really expect to do.

But more recently, other factors have come into play. There is the sheer ubiquity of fund managers, again including hedge and PE funds.

Once upon a time, there was the odd unique hedge fund or unique PE fund such as KKR, but there are now innumerable both mainstream and alternative managers chasing yield in a very low-yield environment. This causes a peculiar statistical problem for any investor when choosing a fund manager.

Because there are so many fund managers from which to choose, including even alternative managers, how can we ever get sound empirical evidence that one fund is outperforming another? For example, if a hedge fund manager purports to have even 10 years of superior returns or apparent alpha, what does that even mean in this day and age?

As is commonly noted, if 5,000 people are tossing coins, some, just by the natural laws of probability, will get 10 heads in a row. Of course, this doesn't make the lucky coin tossers any different from the others or any more likely to get a head next flip around.

Some fund managers may indeed be better performers than others, but the issue is how can one empirically prove that?

There are, it seems, exceptions. The Warren E. Buffetts of the world have produced superior returns across a lifetime, so there are perhaps a cadre of investors who have a unique identifiable gift, but this group is small.

Thus, the modern ubiquity of the industry means that anyone who chooses one fund manager or another, thinking that the manager really can generate sustained "alpha," is probably acting according to the principles of irrational behavioral economics. Investors couldn't really have a hard rational economic reason to select one fund manager over another, so they do it because perhaps they have heard from friends that a given manager is somehow "special" or maybe it becomes something to brag to friends about.

The high fees of alternative funds, the vast number of them and the fact that, on a risk-adjusted basis, very few deliver sustained out-performance, tells us something even more curious. Investing in an expensive alternative fund is a particularly irrational or acute example of Robert Shiller-like behavioral economics in action.

There may also be a small class of artificial intelligence-based fund managers using stochastic and modal systems that do outperform the market such as Renaissance Technologies. By necessity, these can only be few and only manage limited amounts of money or else they become the market and thus distort their own computerized predictions.

So a few of these guys are out there and are quite likely to be producing real alpha and are hence worth investing with. But again, remember that such systems aren't human, and the original question was why do human fund managers still exist?

What about the sales process? Well, selling and establishing brokerage investment accounts and even loan requests is increasingly done online, so there is very little need for the traditional fund management salespeople.

Given all the above, the most rational thing for typical investors to do with their money if they don't want to manage it themselves is to invest it an index tracker fund online. At least the investor will get the market return at a low cost.

But of course, if any type of fund management is easy to automate, it is index trackers, which require relatively simple algorithms to allow for a portfolio to follow a given index. Indeed, most of these trackers are already fully automated.

So unless one is subject to the irrationality of behavioral economics, which many of us are, it is hard to see why, with a possible few exceptions, one needs to ever use human fund managers.

This isn't a call for the end of the fund management industry as money must be invested. On the contrary, it is a call for industry leaders to automate or continue to automate the whole industry in a sophisticated fashion to minimize costs and maximize returns, a call many of them have already heard.

It happened with bank tellers when ATMs were introduced, it happened to back-office processing, and it happened to the trading of fungible assets such as commodities and vanilla bonds. And it is happening to both the back, middle and front ends of the fund management industry, both mainstream and alternative.

It looks likely that this transformation will be largely completed over the next five to 10 years.

This article is commentary by an independent contributor. At the time of publication, the author held no positions in the stocks mentioned.

Jeremy Josse is a partner at Brock Capital, consultant with maxos.ai and the author of Dinosaur Derivatives and Other Trades, an alternative take on financial philosophy and theory (published by Wiley). He has spent more than 20 years in the financial services industry with firms including KPMG, Schroders, Citigroup and Rothschild.

Josse is also a visiting researcher in finance at Sy Syms business school in New York.