Algos: Still Running the Show

JPMorgan estimates that “fundamental discretionary traders” account for only about 10 percent of trading volume in stocks.  The rest?  Passive, semi-passive and quantitative techniques — whose market share has doubled in the past 10 years.

Techniques vary quite a lot.  But, the basic premise is to find factors which are predictive of market direction and invest accordingly.  Factors generally fall into two camps: fundamental (e.g. dividends, ROE, book-to-price) or technical (e.g. volatility, moving averages, momentum.)

Once a factor is widely accepted as strongly predictive of future prices, the vast majority of daily volume is tied to its every tick.  Such is the case with VIX.   Fundamental investors know that rising VIX is a sign of increasing volatility and that they should rein in their exposure lest they suffer losses in a falling market.

Falling VIX, on the other hand, signals the opposite.  If risk is falling, then investors should increase their exposure to profit from a rising market.

A quantitative shop might employ hundreds or thousands of factors in the algorithms with which they trade and invest.  But, over the past several years, one of the most important factors is VIX — which studies have shown is easily and frequently manipulated.

Yet, if you’re watching VIX like a hawk, as do virtually all algorithms, you don’t really care why VIX might be plunging.  You only note that it is; and, you buy rather than sell.

We’ve looked at the longer-term patterns of VIX, which after years of bouncing off the bottom of a huge channel once a year, suddenly spent most of 2017 repeatedly dipping below the bottom of that channel.  To carbon-based factor diviners such as myself, it was apparent that the dips weren’t always in sync with headlines.It quickly became apparent that the dips were often completely out of touch with reality — a plunge in VIX after particularly negative economic data, for instance.  Why would traders short VIX, for instance, in the middle of a plunge in equity futures?

Fundamental stock-pickers — remember, 10% of daily volume — could go all year without worrying about such things.  But, the algos notice immediately and go on a buying spree whenever VIX plunges below important support such as the yellow channel bottom.

And, it doesn’t take much insight to see that, more often than not, such VIX plunges occur when stocks need rescuing or need support in breaking above important resistance.

Consider the DJIA’s chart below.  After reaching its 2.618 Fibonacci extension at 26702, it quite naturally reversed to the next lowest Fib level — tagging the 2.24 at 23824 on Feb 9.

It bounced as expected, but the bounce failed and the index fell to its 200-DMA (the thick red line) which, by then, had risen above the 2.24 Fib.  DJIA formed a rising wedge (a bearish pattern) and fell, tagging the 200-DMA once again on May 3.

This was a particularly weak performance.  So far, May was living up to it s reputation (sell in May and go away.) A breakout was needed.The following day was critical.  Futures had dropped below their 200-DMA overnight — the second day in a row — and the index seemed likely to break down [see: We’ve Seen This Movie Before.]

VIX came to the rescue, plunging sharply in the pre-market.  The drop lasted all of two minutes; but, the algos didn’t care. All they knew is that VIX had plunged 42% from the previous day’s high, had dropped below its 200-DMA, and has even dropped below the yellow channel bottom.  S&P futures bounced 57 points off their overnight low.

DJI tacked on 554 points from its low.  But, it wasn’t out of danger.  Two sessions later, on May 8, it was still struggling to retake the trend line (red, dashed) from its January high.  Again, it needed help.

On May 9 [see: PPI Spells Trouble] we noted that VIX was testing horizontal support, with the 200-DMA just below.  VIX dropped through the support and the moving average, shedding 35% before it finally bounced.

Algos got the message.  DJIA broke out (the yellow arrow), as did every other major index.DJIA ran into trouble again earlier this week, when it seemed stuck beneath its 100-DMA (the yellow line.) Again, all it took was VIX dropping through its 200-DMA. As we pointed out yesterday, it made several 1-minute plunges in the pre-market in order to ensure stocks got off to a strong start.  DJIA gapped higher, tacking on 347 points.  SPX added 23.

Just to make sure it stuck (futures had started to break down from a rising wedge) VIX plunged below the yellow channel line again and broke down — the opposite of the usual outcome — from its falling wedge.

Now, it could be a coincidence that these very significant plunges in VIX occur precisely when equity indices need a boost. Or, it could be that central bankers, their proxies, or players with a stake in the game are routinely manipulating VIX.

Seven years ago, when I began this website, I was often chastised for suggesting QE was artificially suppressing interest rates and supporting stocks.  Along the way, it became apparent that currency (especially USDJPY) and commodity (especially oil/gas) prices were also being used to drive equity prices higher and to produce economic data that wouldn’t upset markets.

The market cap of the S&P 500 is $24 trillion.  The trading volume in VIX futures yesterday was about $240 million.  Which would be easier to manipulate?  Again, a sizeable share of daily trading volume cares only what VIX does, not why.

 

 

 

 


Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.