Tagged: correlation

Correlation

“Why is our stock underperforming the peer group?”

Ever got that question from your CFO or CEO?

We hear it too. Speaking of questions, if you’re in Atlanta come throw them at the panel on High Frequency Trading for NIRI’s chapter breakfast, Friday Nov 16 at Maggiano’s in Buckhead. I’ll be there, and even better, so will the Nasdaq’s Jay Heller and the NYSE’s Rich Barry. We believe our story will rival this David Petraeus thing.

Just kidding. Back to the start, the answer most times lies in fine shades of difference. If program trading for funds and models in your stock is just 1% lower than the average in your peer group, spread over a month you might trail it by 10%.

The good news for IR folks, it’s not because you’re falling down on the job. It’s just math – which you might shake by targeting potential holders with shorter horizons and more aggressive trading proclivities.

Left unattended, these patterns tend to intensify. Mathematical models, identifying discrepancy, reweight allocations and the pattern reverses. It then repeats in shrinking cycles until something rattles the whole market and things reset.

Speaking of resets, high correlation, or uniform trading patterns, is a harbinger of looming group stampedes. Now, don’t worry. We don’t hear a thundering herd. We just see curious and repeating correlations.

Correlation in US stocks hit a record 88% in 1987 during the market crash. Before it happened, correlation soared to 81%, about the same level we saw in 2008 when Lehman left for the big brokerage in the sky and markets went the other direction. (more…)

We were in San Francisco Sunday escaping the heat parching most of the country. Cool heads are better than hot heads, we thought. It was nice to need a sweater.

Speaking of needing things, there’s a flaw in consensus estimates. Consensus by definition means it’s the general view. But the general view reflected by estimates of earnings, revenues or cash flows comprises less than 15% of total market volume.

Across the market, we find that about 12.5% of substantive volume is what we’d call rational – driven by thoughtful investment derived from fundamentals. How can this be? Great swaths of trading today are driven by relative value – the current value of this basket of things versus that basket of things.

Somewhere around 30% of volume is this kind of trading that we consider program trading. It’s driven by market factors and relative value. After all, currencies that denominate securities have only relative and not intrinsic value. Should we not expect trading instruments to behave the same?

What’s more, some 65% of total market volume on average is just air created by the maker/taker model prevailing across global exchanges, in which we’ve all been fed this line of hooey that a massively mediated market is better for buyers and sellers than one with few intermediaries. When in the history of human commerce has it been more efficient to cut the middle man in rather than out? (more…)