Buy high & sell low?

June 19, 2016 03:00 PM

Ask a random group of people on the street what they know about trading in the financial or derivatives markets and they will more than likely reply with the standard “buy low and sell high.” While this age-old saying is technically and partially correct, historically, attempts by investors to pick tops and bottoms have more often than not ended in tears. Markets have always been inherently unstable and subject to the natural forces of fear and greed.

But there is an entire discipline of trading with deep roots of proven success that remove these forces of fear and greed. A strategy that takes the opposite approach to calling tops and bottoms is the standard bromide, “The trend is your friend.”

If you are reading this article, you may already be familiar with trend trading techniques and concepts. But assumptions can be dangerous things. Time and again we have found ourselves left in shock at the deep misunderstanding of core trend trading principles, even by savvy industry professionals, although we probably shouldn’t be.

Market participants have few ways to compare alternative investment returns or investments in managed futures on a timely basis. Most of the benchmark options currently available in the trend- following space involve a simple index of reported returns from select managers, many of whom are blending multiple trading strategies. These measurements can be slow to report and can potentially carry hidden and misleading risks, such assurvivorship bias, selection bias and backfill bias, delivered with a lack of transparency. In an effort to increase awareness of how these strategies behave, as well as create a consistently reliable, useful and understandable index, we began an academic project to publish a transparent and unbiased benchmark index for commodity trading advisors’ (CTA) trend performance we call 40-in-20-out.

Traders should monitor multiple strategies in parallel, even if they are not actually trading them. Understanding the behavior of different strategy sets and how to use them can provide useful insight. As this is something we do for ourselves, we know making this information public would benefit others.

There are scores of books, research papers, podcasts and easily searchable trading rules devoted to trend-following, and we enthusiastically encourage you to investigate the topic on your own.  Thorough research breeds confidence.  However, absolutely nothing will replace living through the inevitable highs and lows of running such a model with your own capital at risk to fully understand the discipline. You can’t tell until you bet.

We could spend more time than the space here allows to discuss the key concepts of trend-following and its intricate variations, but it’s best to start with a simple model: The 40-in-20-out trend index. This is not investment advice, it’s a simple model designed to teach the basic concept of trend. 

What is a trend?

Very simply, an uptrend can be defined as a time series of reported prices with higher highs and higher lows remaining in place for a period of time (see “Positive trend,” below). A downtrend would be the mirror opposite as the market makes lower lows and lower highs.

It is natural to want to be right in life. Although in speculation, being right is overrated, while being stubborn is a quick road to ruin. You can have fewer winners than losers, just as long as the losses are kept small and the winners large. Cutting losses quickly is a key component to the long-term success of a trend-following model (see “Let it ride,” below). 

Famed investor George Soros once said: “It’s not how right or how wrong you are that matters but how much money you make when right and how much you do not lose when wrong.” 

Most successful trend-following managers will tell you the secret to their success is not in their entry signals but in their risk management and bet sizing.                                                                                                               

The 40-in-20-out model monitors and records trade signals in real time for 29 active and liquid domestic futures markets across eight market sectors (see “Trading roster,” right). 

Trade initiation

In a narrow market, when prices are not getting anywhere to speak of but move within a narrow range, there is no sense in trying to anticipate what the next big movement is going to be – up or down.
— Jesse Livermore

Like a spring being wound up, the longer a market remains in a trading range, the more the potential energy becomes stored. At some point, markets will either exceed the previously made highs, or take out the previously made lows. When this occurs there is the potential for a sustained trend.

If one were to look at a very short window, say five days, then there would be a considerable number of observations.  But if one were to use a lookback window of very long length, such as 200 days, then considerably fewer instances are found. If most breakouts fail, would you prefer quantity or quality?

In the 40-in-20-out model we are using a lookback window of 40 days for initiation of trade, which is not too tight and not too loose — it’s perfect for establishing a comparable benchmark for trend trading performance. Positions are initiated the moment the instrument crosses the 40-day threshold in either direction 
(see “Go time,” below).

The Average True Range is a widely available indicator offered on most charting program and platforms. This value represents the range we can expect an instrument to trade from daily high to low on a typical day. Some days the trading range will be much more than the average, some days it will be much less. The length of the lookback window used is important. If you are only looking back over five days, this indicator will respond rather swiftly to changing market conditions, but will also be quick to forget moves from a couple of weeks ago. For purposes of our benchmark model, we are using a 20-day lookback period for the ATR calculation, which essentially captures the past month of trading activity.

This key value in the 40-in-20-out trading model is recalculated daily and used in determining the initial protective stop on trades and calculating volatility-adjusted unit sizing.

If you can’t take the heat from a position moving at least one “average day” against you, you probably shouldn’t enter. 

The disciplined speculator knows exactly how much he is willing to risk on a trade before he even takes it. In this case, we are risking 50 basis points or 0.50% on each trade. Using a theoretical $5,000,000 account size (a big number for most, and a small number for many) this works out to $25,000 of risk on every trade.

The 0.50% value remains constant in our model to avoid any bias. The “risk budget” is then applied against the dollar value of the 20-day ATR for each instrument to compute the number of contracts to trade.  

“Risk calculator,” (below), breaks down trade size by risk. If the ATR is 1.20 points for U.S. Treasuries, and each full point is $1,000, then each signal taken needs to risk $1,200:  $25,000 / $1,200 = 20.8 contracts per signal, which is rounded to the lowest full number. If the ATR in gold is 19.3, or $1,930, we trade 12 contract based on an initial risk budget of $25,000 per trade: $25,000 / $1,930 = 12.95 = 12 contracts.

Trade liquidation

Once a long trade has been initiated, a protective stop is placed at the entry price minus (1 * 20-day ATR). If the instrument subsequently trades below this protective stop price (vice versa for a short position) then it is exited immediately at a loss.

Thus far, entering into trades has been fairly straightforward. And hopefully by now you can see the wisdom in getting out of losing trades in a timely fashion. This then leaves us with one of the most difficult parts of any program — how to handle winning trades.

There is one school of thought suggesting that traders should “pyramid” or add on to winning trades as they develop, each time advancing their protective stop. Another school of thought says a trader should enter into his entire position upon the entry signal, and then reduce exposure, booking profits as the winning position advances. Strong cases can be made for both approaches based on the particulars of the entry metrics.  

For purposes of this benchmark project we are neither increasing exposure as trades develop nor taking profit along the way. Instead, we are using the 20-day lowest low as a dynamic trailing stop for long positions and the 20-day highest high as a trailing stop for short positions.  Recall our original definition of an uptrend as an instrument making higher highs and higher lows. Will the 20-day dynamic stop be wide enough to keep you in the trade?

“Riding the wave” (below) shows a July 2014 short position in crude oil. This trade lasted 250 days and was rolled seven times in the process with a profit of about 34 times the initial risk on the position.

Very few traders out there would have had the foresight to have seen such a move coming, and fewer still would be able to maintain a full core short position from so long ago. In this case, the old school model of a trailing stop based on chart structure — letting your profits run — and resisting the impulse to take profits as the trade moved in the model’s favor worked out very well. 


Some trend-followers prefer moving average systems to chart structure and breakout systems. Some increase exposure as trade develops and others do not. Some have short lookback windows and very tight stops, others take a much longer view and have very wide stops. 

Often the simple solution is the best solution, but there are no perfect answers. 
If you know that going in and understand that the real battle is you against yourself, you’ve got a chance at participating in some big trends when they happen.

Ultimately, the model did “buy low” and “sell high” in the crude oil market in this example — just not at all in the way most people would suspect. Only with the beneficial perspective of hindsight do such moves appear obvious.

The 40-in-20-out system is run in real-time. You can follow along with the project at The 40-in-20-out trading rules are posted posted at

About the Author

Andrew Strasman is a 25-year veteran trader, registered CTA and the sole principal of Totem Asset Group, LLC. He co-founded Transparent Index Group, LLC in 2014 as a platform to build out this and other evolving academic projects of interest in the trading community.