Managed Futures Today Magazine

Controlling risk in a managed futures program takes many forms — everything from properly selecting markets to allocating capital and sizing positions. But limiting losses is also accomplished through the same tool frequently used by individual investors — the stop order.

STOP ORDER BASICS
There are several types of stop orders, but the “stop-loss” order is the specific tool used to limit an open position’s loss or protect its existing profits. For a long position, a stop-loss order to sell is entered below the current market price; for a short position, a stop-loss order to buy (cover) is entered above the current market price. For example, a fund manager who goes long crude oil futures at $88/barrel might enter a stop-loss order to sell his position at $81.20 if a move to that level negates the bullish outlook that triggered the trade. Similarly, a manager who establishes a short position in the 10-year T-note futures (TY) at 128-14/32 might enter a stop-loss order to cover the trade at a loss if price rises to 131-22/32.

A stop-loss order is “activated” only when price trades at the stop-price level, at which point it becomes a market order that will be filled at the first available opportunity. There is no guarantee a stop order will be filled at the stop price, since during very fast, volatile conditions, a market may move a considerable distance before an offsetting order can be found. However, the more liquid the market, the better the odds the order will be filled at or near the actual stop price.

One way fund managers can attempt to reduce this drawback is to use a “stop-limit” order, which functions the same way as a stop-loss order except a limit order (rather than a market order) is activated when the stop price is hit. Stop-limit orders consist of two prices — the trigger (or “activation”) price and the stop-out (limit) price. Returning to the long crude oil example, a stop-limit order could be entered to sell 20 contracts of October crude (CLV11) at $81.20, with an $80.75 limit. The first price is the activation price (i.e., the price to which the market must drop in order at to make the order “live”), and the second price is the limit price that represents the worst price at which the order can be filled. In this situation, as soon as the market trades at $81.20, an order is triggered to immediately sell 20 October crude oil futures contracts at a price no worse than $80.50.

The downside to stop-limit orders is there is no guarantee the order will be filled — partially or at all. If price rapidly falls to $80.50 and continues to decline, a trader may be stuck with an open (and losing) position. (A trader can choose to make the activation and stop prices identical, although in fast-market conditions, this will increase the odds the order will not be filled.)

INTEGRATION OF STOP LOSSES IN A TRADING PROGRAM
Depending on the trading program, stop-losses can be employed in distinct ways — or not at all. One program may use an initial stop-loss to protect a position against losses immediately after it is established; others may implement a stop only after a certain amount of time has passed or a certain price move has occurred. Still others may forego formal stops altogether because their trading approaches (such as some long-term trend-following techniques) simply switch directions (from long to short or short to long) whenever the market reverses direction by a certain amount, an approach commonly referred to as “stop and reverse” (SAR).

Different techniques, including current market volatility, can be used to place stops — and adjust them — in such a way that they reflect changing market conditions as a position progresses. A “trailing stop” is a dynamic stop-loss order that (for a long position) rises as the market rallies or (for a short position) falls as the market declines, protecting open profits.

TRAILING STOP EXAMPLE
Figure 1 shows the application of a hypothetical initial stop-loss and subsequent trailing stop in gold futures (GC). Gold has been in a major bull market for a decade, and let’s assume a fund manager was waiting for an opportunity to enter the uptrend in spring of this year. After a big run-up in April, the early-May correction offered the chance to get in the market at a lower price level, in anticipation of another upswing.

Let’s say the manager’s system established a long position at $1,483 on the close on May 5 (the big down day when price dropped to $1,468, more than $100 below the high price just three days earlier), with the expectation of participating in another major up leg in the bull trend. In this case, the manager’s analysis might have indicated the May 5 decline marked a significant swing low that might be challenged but shouldn’t be penetrated to a significant degree. As a result, the manager might place an initial stop-loss or stop-limit sell order around $1,444 ($24, or 1.6 percent, below the May 5 low), a level that gives the market room to test the “round-number” price of $1,450, which also happened to be a resistance/support level tested by the March highs and the early-April swing low. For the manager, a decisive move below this level would negate the bullish outlook and necessitate getting out of the long position.

In this case, the $24 decline below the May 5 low also happens to represent the typical daily volatility over the preceding 10 days — an idea that will be used to implement a trailing stop that will adjust automatically to changing market conditions. In this case, the daily range calculation used to define the market’s recent volatility is the “true range,” which is the largest of:

1. today’s high-low range;
2. the absolute value of today’s high minus yesterday’s close;
3. the absolute value of today’s low minus yesterday’s close.

The true range more accurately reflects period-to-period volatility in the high-low range because it incorporates the gaps that may occur between bars (i.e., when the current bar’s low is above the previous bar’s high, or the current bar’s high is below the previous bar’s low). The average true range (ATR), which is the average of the individual true range calculations over a given period (for example, 10 or 20 days), is a common measure of market volatility. In this case, the manager is using a comparable gauge, the median true range (MTR), which simply substitutes the median for the average in the ATR calculation because the median is less susceptible to distortion from anomalous large readings (for more information on this point, see “Average and median”). The MTR will be used to adjust the stop to the market’s recent fluctuations, moving it farther away from the current price when volatility increases and moving it closer to the current price when volatility declines.

The manager’s strategy keeps the initial stop in place until the market makes a new 40-day high (a daily high above the previous 40 days’ highs), at which point a trailing stop formula will kick in that will automatically raise the stop-loss level if the market continues to rally. On any day the high is above the highest high of the past 40 days, the stop-loss is raised to:

1. today’s close minus two times the 10-day MTR;
2. the stop cannot be lowered; it can only be raised. If today’s projected stop level is lower than the existing stop level, keep existing stop level intact.

For example, let’s say the market made a new 40-day high today and closed at $1,563.30. If the current 10-day median true range is $17.85, the stop-loss would be placed two times this amount below the closing price:

Adjusted stop-loss level = $1,563.30 — (2*$17.85) = 1,533.40

The logic behind this trailing-stop adjustment is very simple: The manager has determined a down move more than twice the size of the typical daily volatility over the past 10 days (i.e., the 10-day MTR value) means further declines are likely to occur and the long position should be exited. (Note: These values are not based on conditions derived from the gold market; they are simply used to illustrate the basic principles and logic of using this type of stop-loss technique.) Tying the stop level to the 10-day MTR is a way to give the market “more room to breathe” when conditions are more volatile, and to keep the stop tighter when conditions are quieter. Adjusting the stop only when price makes a new 40-day high ensures it is being raised only when the market is advancing.

In Figure 1 the blue down arrows mark new 40-day highs, which represent the days the stop can be raised. The chart shows the first new 40-day high (after the trade entry) more than two months later, on July 12. At that point — and at every subsequent bar with a blue arrow — the stop was raised to the most recent close minus the current 10-day MTR (red horizontal dashes). Over the next month, this formula raised the stop 12 times, from $1,533.40 after July 12 to around $1,722 after Aug. 10. The stop always followed the market higher, and it moved relatively closer to the current closing price when the short-term volatility was lower (mid- to late July), and moved farther away from the current close when the short-term volatility was higher (mid-August).
    
JUST ONE PIECE OF THE PUZZLE
Stop-losses are merely one potential component of the risk-control regimen employed by the typical managed futures program. The preceding example does not reflect other common trade-exit tools that might be incorporated in a managed futures strategy, including specific profit targets or offsetting trade signals from another trade strategy or system. The example simply shows the progressive exit levels that would have been in place for the long trade in the absence of any other strategy components.

In addition to dynamic position sizing, market diversification, the application of multiple strategies operating on multiple time frames, and portfolio equity risk limits, stop-losses can be an effective way to limit risk in a trading program.