How Algos are Killing Off Traders

I’ve been a big fan of chart patterns, harmonics and technical analysis for many years.  They’ve worked well for me, for those who follow this site, and for many generations of traders who came before us.

Since late 2013, however, algorithmic trading has practically ruined these time-tested tools.  Unless you follow the market throughout the day, tick by tick, you might never realize the extent by which they determine stock prices.  I’ve had many conversations with veteran investors who have a hard time believing it until they set up multiple screens and watch the instruments involved side by side.

There are many different types of algos involving all types of instruments, currencies, derivatives, etc.  All involve computers, and many involve high-frequency trading including quote-stuffing (hundreds or even thousands of orders are entered and cancelled within seconds in order to influence price direction), front-running (needs no explanation), layering (placing multiple large orders slightly away from the market to push it one direction or the other) and subpennying (providing a slightly better bid or offer in order to nudge the underlying higher or lower.)  There are many others, as illustrated in this chart from BlackRock (which uses less pejorative terms.)


In my experience, the most influential algo going today involves the USDJPY.  Many investors are aware of the dollar-yen carry trade.  The basic premise is that an investor borrows in yen and invest in stocks.  As long as the BOJ continues to trash the yen, it is presumed that the loan will cheaper to repay in the future – a riskless trade.  And, as long as central banks continue to prop up stock prices (directly in the case of Japan, indirectly practically everywhere else), the assumption is that the other side of the trade also riskless – a safe assumption for several years, now [note: the market averages 11 months between 10%+ corrections; it has now been over three years.]

While the dollar-yen carry trade makes sense in a big-picture sense, many are surprised to learn that it is being used by algos to direct even minor moves in the markets throughout the trading day.  Currency markets are almost entirely unregulated and are extremely easy to manipulate.  There’s a nice collection of FT articles HERE, but any Google search for “forex manipulation” will provide plenty of reading material.  And, given the general lack of volume in the markets lately (gee, I wonder why?), the impact can be substantial.

We’ll look at two blatant examples that occurred earlier today.  They’re not the most egregious I’ve seen, but they have the advantage of being fairly obvious to anyone who knows how chart patterns normally work.  First, we’ll look at the e-minis on a 5-minute chart.  Note the strong trend line (blue, dashed) connecting the highs of the day.  Note also that the declining Imoku Cloud guided prices lower throughout the day.  The chart is cut off at around 3:30pm EST, when ES was coming up to test the trend line for what was probably one last time during regular trading hours.


The 50-period moving average had also guided prices lower, with the 100-period line also pitching in.  A strong move higher at 3:30pm would have taken prices up to the trend line, and perhaps the SMA100 (marked as Point A.)

The USDJPY chart looks fairly similar.  It also bears a strong trend line connecting the highs of the day.  One glaring difference is the it broke above its Imoku Cloud.  Though, it appeared to be contained within the cloud as it backtested the SMA50 and the trend line at around 3:30.


In an unrigged market, one might expect both ES and USDJPY to turn lower in the last half-hour of trading.  From a harmonic standpoint, ES was aiming for a very mild 38.2% retracement (1921.61) of its rise from last week’s low.  It broke through the .236 (1929) without too much fanfare, and was backtesting it at 3:30 (shown below.)

But, the USDJPY reversed higher when ES arrived at 1923.50.  And, when ES reached its trend line and the bottom of the Imoku Cloud for what should have been the start of another downturn, USDJPY suddenly spurted higher – breaking out of its downtrend (the top of its cloud, its trend line, a fan line, and the SMA50) on absolutely no news.   See Point A below.


The impact on ES was powerful.  Within a few minutes, it followed suit – pushing up through the bottom of the cloud, the 100-period moving average, and its trend line.  Again, no news…just a strong signal from USDJPY that it should move higher.  Just in case the signal wasn’t clear enough, it was accompanied by the sudden cancellation of large sell orders that were positioned a few ticks higher, a vicious monke- hammering of VIX and 10-year note futures, and a 25-pt gap higher on Nikkei futures on negligible volume.


The result: ES shot up through the falling cloud, where it lingered until the 10-period moving average (thin, red) could catch up and provide support.  Note that the SMA5 carried it up to the SMA200 (thick, red) where it lingered until (need we ask?) another spurt by USDJPY got it over the hump (Point B.)  I don’t know where it goes from here, as USDJPY is bumping up against its 200-day moving average (102.34.)  It’s not that the market makers and/or BOJ can’t push it on through.  But, for such major moves, they usually wait until the middle of the night after Tokyo is done for the day (4am ET.)

More typically, they’ll bring USDJPY back down to earth while propping ES up directly.  The overnight volume is ridiculously low, and the afore-mentioned tricks will usually suffice without having to involve USDJPY.  It shouldn’t be too difficult, given that ES is now perched safely on top of the rising Imoku Cloud.


The chart below shows both instruments and the various turning points USDJPY provided during the day.

2014-08-12-ES v USDJPY


Again, there’s nothing like a real-time, side-by-side comparison to appreciate the actual impact.  After seeing this done nearly every day for many, many months, it’s become fairly depressing.  The talking heads lament the markets’ low volume, but rarely acknowledge HFT and algos, much less discuss the huge impact they’re having.  The big HFT firms have wised up a bit, especially after Michael Lewis’ Flash Boys exposed many of their dirty little secrets.  They’re careful which way they push the market.

What’s the big deal, as long as HFT and algos are used to drive the market higher?  Who doesn’t want higher stock prices, even if it comes at the expense of the legitimacy of the market?  The regulators are certainly aware of what’s going on.  But, as one client asked earlier today, how can we expect the feds to do something about it, when The Fed is one of the primary enablers?

No doubt, the scourge will continue on just as long as it drives prices higher and provides plenty of commissions to the exchanges – no matter how many traders are run over in the process.   Should it cause, say, another flash crash, you can bet your bottom dollar there will be a swift and thorough investigation.





How Algos are Killing Off Traders — 1 Comment