contrarian Archives - Algorithmic and Mechanical Forex Strategies

Yesterday’s musings on an S&P 500 doji strategy led to a general discussion of stocks and market crises. I promised to analyze a price-moving average cross strategy and to analyze the performance in times of exceptional volatility.

The results are in and they’re exactly what I predicted. I’m shamelessly tooting my own horn on this one – it’s so rare where strategies do exactly what I predicts.

The direction of the returns matches any traders definition of crisis and regular trading periods over the past decade

SPY Crisis Strategy Rules

The trading rules only initiate short trades. No long positions are allowed.

Enter short next bar at market when:
The price crossed and closed below the 20 day SMA on the last closed bar
The trader believes that a crisis environment either currently exists or is about to exist

Exit an open short trade when:
The price crosses and closes above the 20 day SMA on the last closed bar

The position size is equal to a fixed dollar value divided by the current share price. As an example, SPY currently trades at $169.24. If you wanted to control a position size worth $1,000, then the number of shares is the floor of $1,000/$169.24 = 5 shares.

This strategy is intended to be timely for the current trading environment. Based on all of my proposed definitions below, most of the crisis alarm bells are ringing at the moment.

Defining a crisis

The most difficult part of this type of strategy comes from defining a “crisis environment” quantitatively. Crises don’t happen very often by definition, so I don’t think it’s a worthwhile endeavor to try to quantity the crisis bit. That said, a few obvious crisis indicators come to mind based on basic market mechanics.

PE Ratio

The morons on Tout TV (CNBC and company) keep on screaming how cheap stocks are. I’m not a fundamental trader, but the PE ratio contains useful information. Even the most hard core technical analysis buff would agree that companies generating huge positive cash flow and earning growth have to appreciate at some point. The argument isn’t about if that type of stock will rise; it’s just a question of when.

I don’t see how anyone could possibly look at the current PE ratio of 19.3 and argue that stocks are cheap. They aren’t. Stocks are currently very expensive based on a historical comparison.

VIX

VIX is a CBOE benchmark index that allows traders to compare the price of front month options traded on the S&P 500. A more detailed explanation of the VIX is available on Wikipedia if the concept is new. There’s nothing magical about the 20 level. It’s my general experience that most traders consider that number the one to watch. They think of VIX < 20 as "normal" and VIX > 20 as a severe market move.

Most traders regard a VIX above 20 as a dangerous level.

Put-Call Ratio

Options are effectively leveraged bets on market movements with fixed downside risk. When traders load up on puts, they’re expecting the market to fall. When traders buy more calls, they’re expecting the market to rise.

The put call ratio is simply the number of put contracts traded / the number of call contracts traded. A number > 1 means that more puts were purchased that day than calls, indicating an expectation of a market drop.

Theory has it that short term traders are wrong, making the put call ratio a contrarian indicator. I see the put call ratio as more of a lagging indicator.

When a move is real and already happened, traders react too late and buy protection that they no longer need. The 2008 financial crisis a great example when the ratio spiked to 1.5, a wild number. Just in the past week the ratio went as high as 1.3 before settling back down. The volatility in the number indicates a panicky crowd in my opinion.

Margin debt

Leverage is a two way sword. The theory is that it’s a way to multiply returns by risking debt in the market.

Most traders, and especially retail traders, wind up using leverage as the rope to hang themselves with. Stocks are most commonly purchased with cash among investors. Unlike forex and futures where almost every trader enters a position with leverage, the average retail stock trader enters a position using only the cash present in his account.

An increased willingness among traders to move from cash to margin debt is typically a sign of froth, bubble fever or whatever you want to call it. The chart of margin debt from Business Insider and Zero Hedge show that stocks are currently trading near historical highs.

Conclusion

The 20 day SMA price cross strategy is a great way to run account protection whenever market warning signs are going off. The warning signs may not predict the precise market turning point, but the strategy can function as an effective form of insurance.

The strategy would roughly break even over time if someone were foolish enough to run it that way. Say that you mistime the crisis. Big deal. This type of strategy can run for months without causing irreparable harm to the account.

The signals can run in the background. If you’re only a little bit right with your crisis predictions, the risk reward ratio is massively in your favor. If you’re wrong, the consequences appear to be slow losses that lose a couple of percentage points per quarter. If you’re feeling edgy, I think it’s a great strategy to run in the background to calm your mind.

A Donchian channel measures the highs and lows of the price over a certain period in time. A lot of traders use this concept in their trading, although they are not familiar with the name Donchian.

Most Donchian channel expert advisors attempt to catch breakouts. I almost never see people use it with a ranging approach. Most traders want to ride the excitement of an ever-increasing market. The price, especially with the forex majors, often strikes the previous high or low. The price surges for a minute, only to retrace to well within the previous channel.

The hazard of using Donchian channels as breakout strategies is if you jump too early, you risk making a big fuss over nothing. If you jump too late, then you miss the move. I have not found any method for predicting when these moves will happen. My experience with fractal markets is that the period of a new movement, big or small, is totally random. The condensed trading time and low liquidity make it extremely difficult to try catching a move as it happens, at least on an intraday basis.

I have not done any testing on this, but I suspect that a ranging approach might work better. Most momentum traders are weak hands. They only play when there’s action. As soon as the action disappears or reverses itself, they all tend to leave the party. The dominance of retail traders favors a contrarian approach.

Most traders look at similar points to decide when momentum is truly occurring. They use Donchian channels, although different traders tend to use different periods. The important take-away is that the precise price that they care about tends to vary ever-so-slightly based on the period selected. The Donchian price is more or less the same, regardless of the period.

As an example, you might choose a lookback period of 55. The Donchian channel would consist of the highest high that occurred within the past 55 bars. The high could have occurred on the 55th previous bar or 10 bars ago. Time is ignored. The channel’s low corresponds to the lowest low in 55 bars or periods.

Turtle Traders

The most famous Donchian channel method comes from Richard Dennis and his Turtle traders. Dennis and friend argued over whether good traders were made or born. As wildly successful traders, they had several million dollars at their disposal to settle the bet.

The system used the 55 period high and low to determine the entry. When today’s price strikes the highest daily high in the past 55 trading days plus one tick, the trader enters at market. The system focused on commodity futures.

Most people tend to focus on the methodology that they used to pick the market direction. The original turtles argued that their success came from the unique money management and portfolio selection methodology that they used.

As a winning trade increased in value, the Turtle Trader added a second trade to his floating winner. They used recent volatility and their own risk variable, called N, to determine how far or near the second entry should occur from the original. They would do this up to 4 times, eventually letting their massive winners ride for months.

The system worked extremely well through the 1980s. My understanding is that the performance degraded towards the end of the decade.

If you’d like to read through the entire list of the Turtle rules, I suggest that you read through the Turtle Trader PDF that’s been floating around the web for years.