Napoleon famously wanted luck rather than good generals, but new research suggests the two go hand in hand for fund managers.
In a new paper, Roberto Stein of the University of Nebraska-Lincoln examines what are termed ‘lottery stocks’ – risky investments available cheaply with an asymmetric payoff. Previous work by Alok Kumar, now at the University of Miami, established that such shares underperform safer stocks by around 66 basis points a month. Could the professionals beat these odds?
Using data from the Center for Research in Security Prices and the Thomson Financial CDA/Spectrum fund holdings databases between 1980 and 2015, Stein, like Kumar, defined lottery stocks by their low prices, idiosyncratic volatility and idiosyncratic skewness.
A first observation was that a large proportion of US equity funds had at least some exposure to these lottery stocks: from a low of 50% of all funds in the mid-1990s to more than 85% in recent years. However, the majority of these funds allocated very little to such speculative bets. The average fund’s percentage of its equity assets invested in lottery stocks – its ‘lottery score,’ as Stein dubbed it – was consistently below 4%, although for some small-cap and growth mandates it could pass 10%.
No insurance needed
Stein noted that the average lottery stock in his sample made a poor investment, underperforming by 75 basis points a month on a gross return basis or 81 basis points a month on a four-factor adjusted basis, both statistically significant.
Yet the average lottery stock owned by a fund manager was a winner: those held in their portfolios consistently beat all other stocks in the market by 62 basis points per month even in the four-factor alpha model. Funds that dabbled in lottery stocks duly outperformed those that did not by 10 basis points a month, with these excess returns persistent over time.
Investigating those lottery stocks held by fund managers in greater detail, Stein identified several themes. The professionally owned tickets were on average higher in price, lower in volatility and skewness, and almost twice as large by market capitalization – with lower bid-ask spreads too – than the average lottery stock in the full sample. The managers, then, are making relatively safer bets rather than backing the true long shots.
Making their own luck
Stein took his analysis further by wondering whether managers who picked lottery stocks tended to enjoy equivalent good fortune in the remainder of their portfolios.
Among those funds with stakes in lottery stocks, the average only had a 2.7% weighting to that end of the market spectrum. Stein therefore replaced 2.7% of the non-gambling funds’ portfolios with lottery stocks. This boosted their performance: their average monthly return increased from 0.96% before the portfolio adjustment to 0.98% afterwards.
However, that was still below the average monthly return of 1.06% from those funds that did invest in lottery stocks. So of those 10 basis points of excess returns in the lottery funds, only two basis points were attributable to the lucky tickets. In other words, gamblers were better investors than the abstemious even excluding the enhancement from their lottery wins.
‘Given the evidence described above, a valid question would be: where does the rest of the performance difference come from?’ Stein asked. ‘I speculate that non-lottery funds do not invest in lottery stocks because they do not have the skills required to make good picks within these stocks. Moreover, I would argue that the skill shown by lottery fund managers in picking superior lottery stocks is part and parcel with other fund-management skills.’
Stein concluded by mixing his metaphor a little. ‘Fund managers with this talent can be compared with retail investors in the same way as poker players could be compared with roulette players: while in poker luck is still an important factor in determining winners, the player’s skill also plays an important role, unlike the case of roulette and other games where the outcome depends on chance alone.’