Do analysts do a better job of stock picking when the market is down? The last 24 months provide an ideal opportunity to test this notion. In the 12 months between May 2006 and May 2007, the S&P 500 Index rose by 22.8%; in the subsequent 12 months (May 2007 to May 2008), the market declined by -6.7%. If analyst skill depends on the market environment, we should certainly notice a difference when we compare results from these contrasting periods.

A Simple Metric
There are many ways to judge how well analysts pick stocks. One simple metric, illustrated below, is to measure how frequently analyst recommendations outperform the market. ² The chart shows the frequency of outperformance for calls in each ratings category relative to the

S&P 500 Index. If there’s any change in the analysts’ ability to separate outperformers from underperformers in the two periods, it’s not visible in these data. And separating outperformers from underperformers strikes us as the essence of stock picking.

More Complex Metrics Analyst ratings do two main things: they tell us how bullish or bearish the analyst is about his coverage as a whole, and they tell us which stocks the analyst most and least prefers within his coverage. We refer to these elements of performance as coverage timing and stock selection. Both are sources of potential value, but they’re economically and analytically distinct and ought not to be conflated. ³

Timing metrics reflect an analyst’s relative degree of bullishness or bearishness, scaled by the performance of his coverage universe as a whole. An analyst adds timing value by being net bullish (i.e., having more Buys than Sells) when his universe appreciates and net bearish when his universe declines. Selection metrics, on the other hand, reflect the degree to which an analyst’s Buys outperform his coverage and the degree to which his Sells underperform. It’s entirely possible for an analyst to be right on a timing call and wrong about the ordering of stocks within his coverage. For example, the analyst might be net bullish while his universe appreciates – but his Buys might still underperform his Holds and Sells.

To borrow terms from capital market theory (and to use them in an admittedly-imprecise way), stock selection measures how well an analyst generates alpha, while coverage timing measures how well he manages beta.

In our experience, all sell side and most independent research firms are consistently bullish. This means that they’ll add value by timing when their universes appreciate, but not otherwise. That’s borne out by the experience of the last 24 months, as the chart above illustrates. In the first year, when the market was strong, the analysts’ bullish tilt paid off, and the value added by coverage timing was substantial. In the second year, the bullish tilt was still there, but the payoff wasn’t. In contrast, stock selection was a source of value added in both periods.

This shouldn’t be a surprise. It’s reasonable to expect analysts to rank stocks within their coverage expertise, and there’s no reason to suppose that their ability to order stocks correctly will depend on whether the average stock is rising or falling. It strikes us as much less reasonable to expect stock analysts to make a correct call on the direction of the overall market. We should, in other words, expect value added by stock selection to be sustainable, and value added by coverage timing to be random.

Not a Timer's Market...
Rather than saying that the last 12 months’ market weakness produced a stock picker’s market, it would be more accurate to say that it eliminated a coverage timer’s market. Even more precisely, we might say that research providers who were consistently bullish (i.e., most of them) have suffered from their bullishness.

We’ve said that timing and selection are different skills, which need to be measured in different ways, and it’s important to keep those measurements distinct. There may not be stock picker’s markets. But there are stock picker’s metrics – that is, metrics that will help us discern an analyst’s ability to separate winners from losers, without being confused by the market’s directional trend. Understanding this is the key to identifying analysts who can add value on a sustainable basis.

¹ See, e.g., “Though Stock Pickers Struggle, Tech Shares Revive,” The Wall Street Journal, 31 May 2008.

² The data presented here are drawn from a composite of Abacus brokerage clients.

³ See “The Architecture of Analyst Performance” (2006), available at www.abacus-analytics.com, for details on how these metrics are computed.