FYI | Jul 30 2010
By Greg Peel
Last night one might have expected the Dow Jones Industrial Average not to move too significantly ahead of tonight's US GDP release. As it was, the Dow closed down only 30 points. But prior to the close, the Dow was up 87 points in the morning and down 110 points by midday.
Such volatility might offer opportunities for wily traders but it is very unnerving for longer time horizon investors. Volumes are already thin because many investors are still once bitten, twice shy since the GFC, but it's now two years later and still Wall Street is being whipped around quite mercilessly. Such volatility is hardly going to entice the missing cohort of smaller investors back.
What has many older hands in the market concerned right now is the issue of “correlation”. While we all closely watch indices as market benchmarks, we know that those indices are made up of cyclicals and defensives and high volatility “risky” stocks and low volatility “less risky” stocks, and stocks with “alpha” risk unique to themselves and “beta” risk of general market sentiment. What this means is that on any day, the movement of an index is simply a net result of movements of constituent stocks, both up and down, in varying degrees.
This allows the trader or investor to cherry-pick stocks within the index which they prefer to hold, and ignore others – “stock picking” as it is known. Stock picking serves to reduce market volatility because it dampens the herd mentality. However, the twenty-first century has seen the growth of two new elements within the markets – exchange traded funds (ETFs) and high frequency trading (HFT).
ETFs combine groups of stocks as a mini-index of sorts which in effect offer investors the chance to let someone else worry about the stock picking. Each ETF will have a different purpose – it might be a sector portfolio, or a portfolio of defensive stocks, or growth stocks, or value stocks or whatever. It might be leveraged, or it might be short-side.
Either way, EFTs clump stocks together, and in so doing effectively reduce the number of traded securities from individual stocks to lumps of stocks.
HFT is blink-of-eye computer model-driven turnover which works on using tiny parcels of stocks to exploit tiny price discrepancies. Were you able to slow-mo HFT in the market (it happens in microseconds) activity it would look a lot like a shoal of sardines swimming as a group and then changing direction rapidly and uniformly.
Computers are not sentimental. When it comes to opinions of whether a market is good or bad value they are ambivalent. They have simply been programmed to chase incremental profits by moving a lot faster than humans. And they tend to feed on themselves.
Recent data suggest, as noted by CNBC, that ETFs now account for an average 30% of daily stock volume, a level which has tripled in five years, while HFT can account for anything from 50-75%.
What both these numbers suggest is that stock movements within the broad index on any given day are becoming a lot more correlated. EFTs and HFT are tending to push “stocks” as an asset class into being more of a singular asset. EFTs reduce the amount of divergence between different stocks and HFT exacerbates herd-style movements.
Take this to the nth degree and what we would effectively have is one “stock”, just like there's only one “gold”, rather than a collection of disparate stocks which cancel each other out to a certain degree. Were there only one “stock” to trade to represent all equity investment, then movements in that stock could only be extremely volatile.
There is not one “stock”, but intraday and interday volatility in Wall Street markets at present is evidence not only of thinner volumes (a lot of cash on the sidelines) but lesser diversity of “opinion” driving what volume there is.
Amidst all of this, many investors are still scared to re-enter the market since the “flash crash” in May which to this day has never really been fully explained.
