Algorithmic and Mechanical Forex Strategies | OneStepRemoved

The Argument For Simple Systems

by 6 Comments

As quantitative traders, we are naturally inclined to make our trading systems more complicated than they need to be. Tinkering with adding different components attempting to find slight improvements is just what we do.

While that passionate pursuit of improvement has likely paid dividends in other aspects of our lives, it can sabotage our trading systems. We are generally better off to trade simpler systems with great discipline. Of course, that is easier said than done.

simple systems

There is no need to invent a better mousetrap, the simple systems we already have can work just fine with the right discipline.

There was a post published at Systematic Relative Strength that addressed the benefits of focusing on a simple relative strength system. While the relative strength part could be switched out for any other simple indicator, the principle behind the post is important.

The post starts by addressing the idea that no matter what simple strategy you choose, you have to actually use it:

One of the ongoing difficulties for investors is finding some kind of simple method for investing.  Relative strength is just such a simple method.  Even simple methods, however, have to be applied!

The post continues to argue the importance of simple systems rather than complicated ones:

If you can’t follow a method because it is too complex and if you bail in panic during the first downturn, you’re not going to succeed with any method.

So we want to avoid complex systems because they will be more difficult to follow during downturns. That is a good point, but their next point is even better:

Simple systems are generally more robust than complex systems, and relative strength is about as simple as you can get.  Relative strength is not an optimized system—like most simple systems, it will make plenty of mistakes but its simplicity makes it robust.

Ironically, we want to find a system that makes mistakes. That way we know it has the ability to overcome them. That type of robustness is what gives us confidence that the system will be able to handle a Black Swan event.

Human nature, I think, makes it difficult to follow any system, whether simple or complex, so discipline is also required.  Investors will improve their chances for success with a simple, robust methodology and the discipline to stick with it.

It seems like so many quantitative traders are forcing themselves to try to invent a better mouse trap. Why bother when the ones that we already have can get the job done?

In order to develop a profitable trading strategy, we need to have the discipline to stick to our system when the going gets tough. That can be done easier with a simpler system.

Could An Overnight Edge Enhance Your System?

by Leave a Comment

Markets are capable of changing dramatically while we are asleep. This can be seen in all types of markets for many different reasons.

Many traders avoid ever carrying a position overnight. Others ignore this concept because they believe that the impact is minimal over the long-term. Some traders simply avoid taking positions ahead of key news announcements, which is usually what trigger big overnight moves.

overnight edge

Sometimes markets make significant moves while they are closed for the night. Could we develop a system to capitalize on these overnight edges?

Jeff from Alpha Interface took a different approach to this idea when he wrote a recent post summarizing some research he found concerning large overnight price moves in US stocks. Based on the research Jeff wrote about, there may be an overnight edge in trading these moves.

The research Jeff summarized was a paper by Berkman, Koch, Tuttle, and Zhang. They examined 3,000 US stocks from 1996 through 2008 looking for strong positive overnight moves. Here is what they found according to Jeff:

They found a strong tendency for positive returns during the overnight period followed by reversals during the trading day.

This behavior was driven initially by an opening price that was high relative to intraday prices.

It was concentrated among stocks that had recently attracted the attention of retail investors (typically due to a news announcement), it was more pronounced for stocks that were difficult to value and costly to arbitrage, and it was greater during periods of high overall retail investor sentiment.

So basically, when stocks make large positive moves overnight, they tend to go a bit too far. Jeff’s piece continues my suggesting how traders could take advantage of this:

Their tests were predictive and thus indicated that postponing purchases of these stocks from the open until later in the day can avoid these hidden costs.

Similarly, selling these stocks at the open, rather than later in the day, can lead to major improvements in performance.

Obviously, this idea would need a lot more development before it could ever be a legitimate strategy. However, it is a very interesting concept.

It would be interesting to research how many “significant” overnight moves actually happen over the course of a year. This would probably vary a great deal based on the stock universe and your definition of “significant.” From there, you would want to determine the percentage of those trades that would have been profitable and calculate the returns on both the winners and losers.

While this system would require a lot of effort to develop, it could be the type of short term strategy that keeps your capital out of the market most of the time. That would give it the added benefit of protection from Black Swan events and some slight interest income from the risk-free asset you kept the capital in when there were no trades to be made.

Testing the Best Strategy for Handling Broken Arrow Trades

by Leave a Comment

One of the most important concerns of any trader is protecting capital through risk management. One of the biggest threats to risk management is positions that suddenly crash and lose a large portion of their value overnight.

Anyone who holds trades overnight is likely to experience this sickening situation eventually. Many times, these dramatic and sudden losses will completely take us by surprise. It is important to have a plan in place to handle this possibility so that you are not tempted to make an emotional decision.

broken arrow

While broken arrows are rare, you should still have a plan in place to handle them.

Cesar Alvarez calls these kind of collapses broken arrows because the chart will show an increasing trendline that is broken by a sudden drop in price. He recently wrote an article investigating which exit strategy would be the best way to handle a broken arrow.  He started that piece by addressing the importance of having a plan in place:

If a Broken Arrow happens to you, do you know your exit strategy?  I can tell you that if you hold trades overnight, a Broken Arrow will eventually happen.

When a trade goes Broken Arrow, as it did with EXPE, the first thing I do is go with my plan.  Not a plan that I invent after it happens, the plan I created before I entered the trade.

In this particular case, my plan was to simply hold until the next 5 day rotation period (I was trading this in a rotation system).

In order to test which broken arrow strategy would do the best job of recovering capital, Alvarez needed to identify a number of broken arrow trades. He chose to use only market data from 2009 to the present so that the results would reflect our current bull market situation. Finding enough broken arrows still proved a difficult task:

First I looked to see how many S&P 500 stocks closed down 1 day 25% or more. But only 8 did. I like to see at least 100 trades but prefer at least 200.

Next I tried down 20% or more and this only gave 34 trades. Trying losses of 15% or more gave 121 trades. Enough in a pinch but I really wanted more. So I devise a way to take a look at a bigger universe of stocks.

By expanding his universe to the top 1000 dollar volume stocks, he was able to increase the number of trades to 318. This gave him the universe to test the following three exit strategies:

  1. Exiting the next day on the open
  2. Exiting on the next open when the RSI(2) closes above 70
  3. Exiting five days later on the open

Surprisingly for me, each of the three methods returned better results than exiting at the close of the day the broken arrow occurred. Exiting on the next day produced and average profit of 0.77%. The other two exit strategies did even better:

Looking at the Top 1000 stocks, the results between exiting when RSI(2) > 70 or exiting five days later give about the same results. But more importantly, almost 2x better results than exiting the next day’s open.

Based on the difficult time Alvarez had identifying enough broken arrows to test, this is not the type of situation you will find yourself in regularly. However, it is important that you have a plan in place to handle this type of situation should it occur. If you trade long enough, the odds are that you will eventually be confronted with a broken arrow.

Black Swan Protection – Guarding Against Catastrophic Losses

by 2 Comments

The most critical component of any system has absolutely nothing to do with entry or exit signals. It isn’t trailing stops, portfolio allocation, or position sizing either. All of those more technical and better documented topics mean nothing without a plan to avoid the next Black Swan.

The single most important aspect of any trading system is how well equipped it is to protect your capital against a catastrophic loss. Does your system have a plan to handle such an event?

Guarding Against the Black Swan

Trend following expert Michael Covel has spent the entire year traveling around eastern Asia giving talks in an effort to promote systematic trend following. In a number of his recent podcasts he has discussed one aspect of his talk that really strikes a chord with me personally. He explains that he generally gives these talks to conservative, long only fund managers who aren’t very receptive to his systematic approach.

The audiences are surprised when he suggests that there is a Black Swan circling over our heads. He explains that we have no clue when or where it will land, but we know it will eventually land somewhere. What are we doing to protect ourselves from it? What safeguards have we put in place? Interestingly, the audience members have not put much thought into this possibility.

We know from reading the work of Nassim Taleb that Black Swans occur far more frequently than statistics suggest. Our human nature makes us prone to underestimating both their frequency and severity.

Black swans affect trading systems

The Problem With “Conservative” Investing

My mother and her husband recently sold a rental property and netted about $60,000 on the transaction. They took that money and did what they have always been told they are supposed to do with it. They invested it in index based mutual funds. They have no clue about the risk of ruin that they are exposing themselves to because their “investment guy” told them this was “conservative.”

The problem with my mother’s “conservative” investment is that she has no exit strategy to preserve her capital. She simply plunked down all the money that she had at close to the all-time highs of the general market. Now, her only plan is to hope for the best. If a Black Swan lands in Syria next week, she’ll simply chalk it up to “bad luck” and hope for the market to rebound.

It’s possible that she has made this investment at the beginning of what will become the next great bull market. It is also possible that her index funds will be worth half of their current value by the end of the year. Either way, she has no plan in place to respond to either scenario.

Analyzing How Systems Protect Themselves

If you are reading a website about systematic trading, you obviously have more sense than my poor mother. This means you probably aren’t invested in long term, buy and hold index funds, but have you given enough thought to how your approach will handle a Black Swan scenario? How would it have fared in 1929, 1987, 2001 or 2008?

In addition to profiling a number of different trading systems on this site, I have also been paper trading a few very simple systems on my own blog. One of the most interesting aspects of this experiment has been the ability to see how different strategies react to uncertain market conditions in order to protect capital. My mother’s approach has no such protection.

Some systems, like a lot of the mean reversion systems I have profiled, don’t have any set downside protection. They could theoretically hold a long position all the way down to zero, or hold a short position to infinity. While these types of catastrophic losses have a very low probability of actually happening, Taleb suggests that they are actually more likely than we realize.

Systems like the 10/100 Long SPY Long Only, the 89/13 Breakout Strategy, and the 50 Unit EMA Strategy are all built to exit positions at certain points. While this may seem like an obvious point, it is critically important.

Another black swan is going to occur. We don’t know if it will be because of Syria, Greece, or something totally different that hasn’t made the news yet. We don’t know if it will happen next month, or three years from now. We just know that it is going to happen. We need to make sure that our systems are prepared to handle it.

Antifragile Trading – Enhanced By Volatility

by 2 Comments

In Antifragile, author Nassim Nicholas Taleb presents the argument that the opposite of fragile is not simply something that is unaffected by stress or pressure. He argues that there is a quality called antifragile where something actually improves as a result of that stress or pressure. Taleb categorizes things on a scale from fragile to antifragile, with a mid-point he calls robust.

When this idea is applied to trading, the first comparison that jumps out at us is between buy and hold investors and systematic trend following traders. Systematic trend following traders excel under extremely volatile situations, are virtually ignored by intellectuals in their field and are capable of handling deviations because they are not predictive in nature.

An Illustration of The Antifragile Concept

Taleb illustrates his concepts of fragile, robust, and antifragile through Greek and Roman mythology.

Fragile is represented by the story of Damocles, who was invited to a dinner at which there was a sword hanging over his head that was held up by a horse’s hair. Damocles is considered fragile because at any second the hair could snap dropping the sword onto him.

The concept of robust is represented by the Phoenix, which is a bird that is able to be reborn from its own ashes. No matter how many times you attempt to destroy the Phoenix, it continues to return to its original state, unharmed.

Antifragile is represented by the Greek Hydra, which had multiple heads. Every time one of its heads was chopped off, two more would grow back. Any attempt to hard this creature actually made it grow stronger. This is the core concept of antifragile.

An Application To Trading

One of the characteristics that can help measure the fragility of something is how it responds to volatility. This allows us to easily apply this concept to trading, because volatility is actually measured in most markets.

VIX Monthly shows antifragile periods of strength

A monthly VIX chart dating back to 1985

If we look at a long term chart of the VIX, we can see four distinct areas of extreme volatility: The Market Crash in 1987, The Asian Currency Crisis in 1997-98, The Dot-Com Bubble & Crash from 1998-2002, and the Financial Collapse in 2008. These highly volatile periods were high stress situations that would have made antifragile trading strategies better while making fragile strategies worse.

It is no coincidence that it was during these exact periods that systematic trend following traders posted some of their best years, and at the same time, buy and hold investors lost fortunes.

Systematic trend following strategies are not only built to handle these kinds of high stress, high volatility situations. They become stronger from them simply because they provide traders with a way to change direction. This ability to go with the flow during even the most chaotic times is a stark contrast to the rigidity of a buy and hold strategy where the investor has no choice but to sit and watch their portfolio crumble.

Overcompensation & Overreaction

“Intellectuals tend to focus on negative responses from randomness (fragility) rather than the positive ones (antifragility).” – Taleb

Again here, we can see a clear application of this with respect to trading during high volatility periods. In each of our four high volatility trading periods from the past 25 years, most of the intellectuals and so-called market experts were focused on the negative responses, which were the buy and hold investors who were losing their fortunes left and right.

During each of these periods, there was very little focus on traders who were benefiting from the increased volatility. The antifragile traders seemed to go unnoticed, and we know now that systematic trend following traders where the ones who were benefiting from these volatile markets.

Layers & Deviation

Taleb also discusses the complexities that can arise when looking at fragility through different layers and dimensions. He illustrates that since what is good for the antifragile is bad for the fragile, then what is good for the fragile is likely bad for the antifragile.

We have already discussed the huge success that antifragile trend following strategies have seen during times that were terrible for fragile buy and hold strategies. When the inverse situations occurs where markets have little volatility, we have seen that systematic trend followers and buy and hold investors can produce rather similar results. The key difference though, is that the antifragile systematic trend followers have the ability to adjust when a black swan shows up.

“When you are fragile, you depend on things following the exact planned course, with as little deviation as possible – for deviations are more harmful than helpful.” – Taleb

He continues to explain that the fragile is very predictive and that being so predictive is exactly what causes fragility. This is the problem with buying and holding with the expectation that a position will go up over time. The entire strategy is predictive in nature, and that is exactly why it becomes fragile. It can’t handle a deviation from what it predicts.

On the contrary, a systematic trend following approach can, by definition, handle any deviation from what is expected because it simply goes along with that deviation. This ability to adapt and change is magnified during periods of high volatility, which is when systematic trend following is at its best.