View

Day Trading and Volatility: Looking for a Link

December 19, 2005

 

In a previous newsletter, we explored the notion that daytrading returns are helped by volatility. There are various ways of defining volatility, both technical and non-technical, and we each have in our gut a pretty good idea of just what this term refers to. And it is an important idea, since conventional wisdom has it that day trading system returns are driven mostly by volatility. The idea is that trading systems are built to profit from big moves. It does not matter whether the moves are to the upside or the downside — if the move goes against the automated system its entry stop and its logic will combine to get it out of the trade quickly (if it is a well-written system) and if the move goes in its favor it will ride the trend as long as possible. Although no system will win every trade, over many trades a successful system will make money. The important point, though, is that in a more "volatile" market, there should be more opportunities for a system to profit.

The most popular shorthand measures of volatility are probably the ones used in the newsletter mentioned above, the VIX and ATR. Just to recap quickly, the VIX is an index developed by the Chicago Board Options Exchange to use option volume and "implied" volatility to produce a measure of investor sentiment regarding volatility in U.S. equity markets for the upcoming 30 days. ATR (Average True Range) is a simple technical indicator that measures the difference between the high and low prices for a time period. Both the VIX and ATR do a good job measuring what they are designed to measure — investor sentiment and the range of prices seen in a given trading period. But these measures of volatility do not show much correlation with the day trading returns we have seen at Attain. The VIX 3- and 6-month moving averages and the 6-month ATR show a correlation of around .5 with a portfolio of systems, but these slow-moving indicators describe large market movements lasting days or weeks, and they are designed to measure how much a market moves or is expected to move, not the manner in which it moves. Specifically, we would like to separate out choppy markets from trending markets as well as measure volatility in absolute terms.

The ADX indicator can help us achieve this goal. ADX (the Average Directional Index) was developed by the influential technical analyst J. Welles Wilder in his book, New Concepts in Technical Trading Systems, published in 1978. The purpose of ADX is to measure the "trendiness" of a market, giving an indication of whether moving-average-crossover signals should be followed (in high-ADX conditions) or ignored (in low-ADX, "choppy" conditions). ADX is scored from 0 to 100 and in general, a value below 20 indicates a non-trending market and above 40 is a trending market.

Because we are concerned with intraday trends of the sort that would be of use to day trading systems, we want to look not at ADX applied not to daily bars as in our previous study, but rather on some shorter time frame. Four arbitrary time frames were chosen for this study: 5 minute bars with a 60 minute moving average, 5 minute bars with a 30 minute moving average, 60 minute bars with a 360 minute moving average, and 30 minute bars with a 120 minute moving average. These values were obtained via a custom EasyLanguage indicator for Tradestation that was applied to a chart of the back-adjusted SP contract, going back to January, 2000. The ADX value for each bar was output to a file, and then the average ADX value for each month was determined.

To test our theory that ADX values correlate positively with day trading returns, we found the correlation between monthly average ADX values and monthly returns for Compass SP and a $100,000 portfolio of day trading systems. This portfolio was intended to give a better gauge of the performance across all day trading systems traded here at Attain, and included Compass SP (1 contract), RMESA5 SP (1 contract), Impetus ERL (2 contracts), and RC Success ES and ERL (1 contract each). Actual client fills were used in the study wherever possible, and for dates where actual results were not available, computer generated fills were used.

The results were disappointing, even more so than the poor correlations found in the previous study. In our matrix of correlations, the only statistically significant relationships we found were between the 30 minute and 60 minute ADX measures and the day trading portfolio. Compass SP itself did not show any relationship at all with the intraday ADX observations, with correlations ranging from 0.015 to 0.112. The 30 minute ADX / day trading portfolio correlation was the most significant one found in the study at 0.374. With 49 months of data, we have a Student's T test statistic value of 2.8, meaning this is definitely a "significant" relationship. But — it is not the b relationship we were looking for. Furthermore, a close examination of the data points shows precisely the opposite of what we would expect if conditions described by the ADX indicator were driving day trading returns: in the months when the ADX value is above its mean, the returns of the day trading portfolio were, on average, below the mean! This can only mean that the positive correlation between ADX and the portfolio's returns was a coincidence driven by the fact that the portfolio has had generally positive returns. Another way of looking at the relationship is using a value of 1 to represent months with above average ADX measures and 0 for below average months. Using this "binary" set of observations we found a correlation of only 0.167 with the day trading portfolio's returns, which is not statistically significant.

Simply stated, the goal of this exercise is to find a way to describe, in statistical terms, the type of market in which day trading systems succeed. Taking Compass SP, for example, it lost nearly $4000 per contract in September 2005, but made over $11,000 per contract the following month. The logic in Compass did not change, so what was different about the market? In our experience, what day trading systems thrive on is intra-day "follow through" combined with a wide trading range. Although ADX seemed like a promising indicator to measure the degree of "follow through" in a market, in this study it did not live up to its promise and our quest for an improved measure of day trading conditions must continue. It may be that ADX is still part of the solution, but it must be used in the correct way. For example, an ATR 10-day moving average could be used to scale ADX, giving more weight to days with a greater trading range. In the meantime, however, the nature of the relationship between volatility and day trading returns remains as elusive as ever — if it exists at all.

- Jeff Eizenberg

IMPORTANT RISK DISCLOSURE
Futures based investments are often complex and can carry the risk of substantial losses. They are intended for sophisticated investors and are not suitable for everyone. The ability to withstand losses and to adhere to a particular trading program in spite of trading losses are material points which can adversely affect investor returns.

Feature   |   Week In Review   |   Chart of the Week   |  

Chart of the Week : ADX and Day Trading Returns

Feature   |   Week In Review   |   Chart of the Week   |  

With a FOMC meeting on Tuesday, a Quadruple Witching on Friday, along with the annual NASDAQ re-balancing also on Friday, traders had plenty of market activity to trade around this past week. But it wasn’t meant to be as US Stock Index trading ranges remained small on moderate volume. For the week SP 500 futures managed a small gain of +0.50%, while NASDAQ futures remained unchanged. There was some aggressive selling in the small caps early on Friday causing Russell 2000 futures to drop -1.18% and Midcap 400 futures to fall -0.56% for the week.

The commodity markets were much more active with the energy and metal sectors enjoying big sell offs while grains, bonds, and soft’s all rallied throughout the week. Bonds were sparked by language in the most recent FOMC announcement that hints the 0.25 point rate hikes might soon be coming to an end. 30 year bond futures gained +1.03% for the week. The same language did not help the US Dollar with the US Dollar Index falling -1.68% on the news. For the most part foreign currencies rallied with Eurocurrency gaining +1.58% and the Swiss Franc gained +0.82%. The Japanese Yen +4.12% had the biggest gains as the Japanese Government is now hinting at interest rate hikes of their own.

Perhaps even more surprising than the language in the FOMC announcement was the drop in energy prices this week despite the very cold temperatures around the US. Rumors of increased scrutiny on Over The Counter (non-regulated) energy trading by the US House of Representatives may have been the cause of the sell off as crude oil prices fell -4.28%, Unleaded Gas dropped -3.27%, Heating Oil traded -1.15% lower, and Natural Gas futures (which were the focus of the US House committee) fell -4.60%.

The high-flying metals markets sobered up as well last week. Prices fell across the board with many traders taking profits on long positions. Palladium led the fall dropping -7.63%, followed by Gold which was down -4.58% and Platinum which lost -4.56%. High Grade Copper was the only metal market not affected by the sell off with prices remaining unchanged for the week.

Finally, tropical’s and grains continue to be bullish with Sugar futures climbing +5.19%, Cotton was up +1.27%, Soybeans jumped +3.97% higher, Corn rallied +1.84%, and Wheat climbed 3.98%.

***Day Trading***

The negative tone on Wall Street continued last week and many investors and analysts alike are wondering whether the “Santa Claus” rally already came and went in November. Regardless of the direction of the market, day trading systems are looking for volatility to pick up some steam heading into the New Year.

The lack of movement in the futures led to losses for the majority of the day trading systems last week. With that being said, the losses were negligible because of the narrow trading ranges.

RC Success ES was the only system to finish above water with a gain of +$115 per contract. Clipper traded three times last week but finished the week right where it started with $0 in profits-nothing gained, nothing lost.

Compass traded twice for a loss of -$75 but recouped some of the loss from earlier in the week with a winning trade on Friday to finish the week on a positive note. Helix ES traded seven times for a loss of -$93.35 per contract. Electric Daybreaker II had trades in the NQ and ES that amounted to losses of -$150 and -$222.50 respectively.

Elsewhere, R-Mesa eRL had one short trade from Thursday for a loss of -$200. Impetus eRL had a similar short trade on Thursday that amounted to losses of -$224 for the week. Sticking with the same theme, Nautilus ES went short on Thursday for a loss of -$242.50.

Daybreaker SP and R-Mesa SP each traded once favoring opposite sides of the market but the final tally was similar: Daybreaker lost -$300 and R-Mesa -$350.

***Swing Trading***

Despite the low volatility week swing trading systems mostly held their own as they worked to identify any would be trend for the week and weeks ahead.

The trade of the week goes to Tzar. Both Tzar eMD and ES seem to have timed the market perfectly as they reversed short last week locking in solid gains. On the week both markets posted open and closed trade gains of +$1,825 and $1,232.50. Tzar is currently also short the eRL earning +$455 and long the NQ from Nov 16th making +$724 including the roll. Tzar has recently come to life after struggling most of the year.

Jumping ahead and also having an impressive week were Mesa Bonds and Notes which capitalized on a rallying bond market. Mesa Bonds gained +$1,156.25 and Mesa Notes gained +$640.625 in open trade equity.

Back in the index markets the rest of the results were mixed. Those gaining ground included Seasonal ST ES +$382.50 and Delphi eRL +$75.70 in open and closed trade profits. Systems on the losing side included Axiom eMD -$170.30, Axiom eRL -$460, Delphi eMD -$502, Axiom NQ -$572.90, Axiom ES -$772.50, Eclipse eRL -$780, Athena eRL -$960, Jaws Narrowneck bond portfolio -$1,087.50, and Axiom CL 90 and 135 which lost -$3330 and -$3,430 respectively.

With respect to the Axiom trades, last week was difficult week as the market jockeyed in both directions but could not find a trend. All this is to point out that tends do emerge out of consolidation and that all 4 index markets are currently short and holding onto sizable gains as the bears came out in force today. SP futures closed down -0.50% and eRL futures closed down -1.2%.

***Long Term***

At the beginning of December long term trend following systems were hoping that the momentum secured in October and November would lead to big gains to close out the year. However, after watching both foreign currencies and metals move against trend followers last week it appears that a profitable scenario is unlikely for this year. But heading into the New Year most trend following systems seem set up to succeed with long positions in markets like Sugar and Cotton, and short positions in both US and Foreign Bonds. Throw in a rally in the grain markets or a big sell off in the energies and trend following systems appear to be set up for a potentially great 1st quarter of trading.

This past week Sugar continued its reign as the top trending commodity of the week. After years of stagnant and choppy market conditions Sugar futures have exploded in 2005 with prices gaining approximately 62% since mid-April! After gaining another +5% last week this rally is not showing any signs of slowing down and could be the catalyst for rally’s in other tropical markets like Cotton, Coffee, and Cocoa. Not to bad for a commodity most restaurants give away for free!

Heading into this week systems with long Sugar positions include: Please Note: Most Sugar trades have been rolled within the last two weeks of trading and because we are required to use actual customer positions we will only mention open trade profits for the most CURRENT leg of a trade:

Aberration Plus is long sugar for gains of +3735.60 per March Sugar contract. The system has also recently entered short in Bean Oil for open trade profits of +$40.00 per contract, and long in the Mexican Peso for a loss of -$925.00 per contract.

SEMA4 Symmetry is also long in sugar for gains of +$4015.60 in the March contract. This system has also found success in Gold for profits of +$500.00 per contract and in KC Wheat for profits of +$687.50 per contract. SEMA4 did give back some profits in the Dollar Index last week and is now losing -$1650 per contract in the March Dollar Index contract.

Axiom LT is the final system long in Sugar and it is long in the London version of the Sugar contract for profits of +$5485.00 per March contract. Axiom LT is also holding long in the London Coffee for profits of +850.00 per contract and in the London Cocoa for a loss of -$280.00 per contract.

Please Login to: http://www.attainaccess.com for the latest updated statistics.

IMPORTANT RISK DISCLOSURE
Futures based investments are often complex and can carry the risk of substantial losses. They are intended for sophisticated investors and are not suitable for everyone. The ability to withstand losses and to adhere to a particular trading program in spite of trading losses are material points which can adversely affect investor returns.

Feature   |   Week In Review   |   Chart of the Week   |