Black Swans

How to turn them into a Golden Goose

Few if any people can predict Black Swans. Yet history suggests it is possible. After all Paul Tudor Jones used the map provided by 1929 to make his name and fortune in the stock market crash of 1987 and the period after. Last month he invoked the threat of another Black Swan.

Such predictions can clearly pay huge dividends, but so can knowing how to react and trade Black Swans once they have landed.

Over the last few weeks we have looked at the different types of news and how markets react to them. Our event and earnings fractals show how reactions to news are often predictable, or at least “they have a habit of moving in a similar fashion before and during such events as this reflects decisions made by traders confronted by similar risk assessments within an underlying narrative or trend.”

Many traders have taken this thought process a step further to conclude that technicals actually lead news or can predict the news itself. While there is some truth in this, the full answer is, as usual, a lot more complex. In this week’s newsletter we will investigate how and why technicals can predict news, and cover instances when technicals have no predictive power (but a great deal of reactionary value) as the news is completely unexpected; the dreaded Black Swan.

Firstly we have to qualify what technical analysis actually does, or why it works. At its core, TA looks at what participants have done and how they have positioned themselves. This often reveals what they will do next. For example, when a head and shoulder pattern forms (showing bearish positioning) and triggers, technicians know is likely to happen next and how participants are likely to react. When combined with news, or news cycles, it can often seem like the pattern predicts the news.

Take the Tesla (TSLA) chart below for instance. The head and shoulders developed in H2 2017, then triggered and re-tested just ahead of earnings. Did it predict earnings would be bad?

You could argue the pattern did in fact predict the news. But rather than say price action “predicted” something, a more accurate description is to say that participants correctly anticipated a particular outcome. Semantics perhaps, but clearly there was heavy selling in anticipation of a bad report, and those sellers had correctly “guessed” the right outcome (and quickly covered on the earnings gap down). This is the most common way for technicals to lead a news event, and can be seen with many scheduled economic and earnings releases. In many cases large players position early in anticipation of a particular outcome and others follow.

There are various technical aspects which can help the probability of anticipating a certain outcome on news. The context for the Tesla (TSLA) chart above was bearish on the higher timeframes as the head and shoulders formed after a large failed break-out and what looked like a complete trend sequence. But for technicals to accurately predict news, with absolutely no guesswork, the news itself must be known to insiders and acted upon. This relies on a leak of information to people with deep enough pockets to move price, and is usually only possible with smaller stocks and specific types of news. Not only this, but the insiders must act honestly and reliably to the news they are privy to. If bad news is coming, they may actually try to construct a fake rally to sell into.

It is hard to therefore conclude that technicals predict news. Yes, they can, and do, but not consistently. At Matrixtrade we accept this and use probabilities rather than absolutes. When a stock shows relative weakness in performance and a weak structure associated with selling, the probabilities shift towards a bearish news flow. We don’t always know what this will be exactly, but someone, somewhere probably does.

A good example of this comes from the Facebook (FB) debacle in March this year. We actually thought that something bearish could be brewing, although we had no idea what it could be as there was no news scheduled. This was our view on March 12th.

As it turns out our hunch was right as the next weekend the Cambridge Analytica scandal broke and the stock plummeted. We can assume quite a few people knew the news before it made headlines and those on the inside had a chance to position for it. Thankfully their footsteps were on the chart.

Yet this is not always the case. When a news or event comes completely out of the blue and participants have absolutely no chance to position for it, technicals can have no predictive power. These true Black Swan events are rare, and while scary, present opportunities as the price reaction (and participant’s actions) are almost always the same.

We won’t cover Black Swan theory (developed by Nassim Nicholas Taleb) in this article, but will just borrow the term. The best way to define a Black Swan is a move of significant proportion that could not reasonably have been predicted. Although there are many moves every day that seem to have no apparent reason, it is the scale and rarity of an unpredictable Black Swan move that justifies its name.

Many traders catch or indeed get caught out by Black Swan moves for reasons that were not apparent at the time. Natural disasters, terrorist attacks, perhaps even a Trump tweet can be grouped together as market moving events that are only reasonably explicable after the event. Some are unfathomable such as the May 2010 Flash Crash whose cause is still open to debate. Either way the first reaction of the market to a Black Swan event is to sell and de-risk, then ask questions and re-assess later. This first move lower is often scary, but once it is underway we can immediately start looking for trading opportunities as technicals—and fractals – take over. Price reactions repeat as human behaviour in times of fear and panic repeat.

9/11 is a classic Black Swan. It shocked the world, and sent panic through the markets. US equities shut down and when they re-opened the S&P500 plunged 13% in the next five sessions. The reversal eventually came from a technical level—the September/October 1998 pivot lows—but it is the subsequent reaction which is interests us the most. Price recovered all the previous decline in 12 sessions.

As we will see, returning to this point of origin is a common occurrence, even when the Black Swan event happens in the midst of a market crash like 2000-2001. It is almost as if the market erases the move, then carries on as if it never happened. Our prediction for the eventual Trump Crash is that it will exactly wipe out the Trump gain as if he was never elected! The 2011 Japanese earthquake and subsequent Fukushima nuclear disaster is another example and had a significant effect on many markets, especially the Nikkei.

Prices crashed 20% in two sessions, but even this move returned to the daily range where the crash originated. This ‘return to the scene of the crime’ is not only a target for any longs, but sets up a trade in the opposite direction. The logic is simple; those holding before the crash take the chance to break even and the selling creates resistance and a fall in price. We don’t know how much of a reaction there will be, but the vast majority of times there is at least some downside move, especially if the underlying trend was lower anyway.

Looking further back, here are some other memorable Black Swan events such as Pearl Harbor in 1941.

and the assassination of President Kennedy in 1963.

You couldn’t really call a tweet from President Trump or a stray comment from a Fed member a Black Swan event, but they do take the market by surprise and the reaction is to sell first and ask questions later. The overall path is very similar to that outlined above, although on varying scales.

Black Swan events are therefore very tradeable in several respects.

Firstly, we may not predict the cause, but we can identify the conditions where one is likely to occur. This cannot be done reliably by technical analysis alone. There is a wealth of evidence—much of it intermarket and fundamental – that says a Black Swan is flying on the horizon on a similar flight path to 1929 and 1987. Most markets back then very usefully produced a first leg down and then an upside correction before the dramatic decline.

There are a few exceptions to this pattern notably the Nasdaq Tech 100 (NDX) in 1987 and Enron (ENRN) in 2001 that could actually help time the turn and increase leverage this time round. We are yet to see any of these conditions, but next week we will be doing a full FITS audit on this comparison (fundamentals, intermarket, technical and sentiment) for subscribers to pinpoint when and how we should prepare and trade it. But the similarity now to the almost unique interaction (correlation) of different markets to 1929 and 1987 remains very clear.

Secondly, and partly dependent on the cause, we can pick the post-crash bottom based on technicals and go long for a clear target to the point of origin.

Thirdly, once this point has been reached we can fade for a downside reaction.

Fourthly, and finally arguably the best opportunity of all, we can buy for the resumption of the longer term trend. We can actually identify all four of these points on a recent chart of the S&P500 (SPX), and whether or not you want to call the February 2nd 2018 NFP a Black Swan, the pattern described throughout this article is very clear.

Defining a Black Swan as it happens in the market is curiously easier it seems than defining it in the classroom.

Just as technicals can predict a move that would suggest a news event is forthcoming, the history of all four FITS (fundamentals, intermarket, technical and sentiment) can forewarn the coming of a Black Swan. They may not be able to predict an exact cause, and as we have seen, technicals are more useful in the aftermath than a stand-alone predictor, but who cares—if we have a map that shows clearly where the Black Swan will fly before during and after, then we have a golden goose.

Here’s to making the very best, and good luck and good trading.


Ed Matts, Founder
and Andrew McElroy, Chief Stocks Analyst
Matrix Trade

To receive our weekly articles by email in advance of publication, please sign up for our Sunday newsletter.