Calling All Value Investors - Investing in Drawdowns
September 6, 2005
After watching the Compass system hit new equity highs 5 days ago, representing a return of nearly $22,000 per contract since hitting its drawdown valley on August 4th, 2004 - it seems as good a time as any to re-visit the logic of investing in systems at or close to new maximum drawdowns.
The basic logic of investing in a system in a drawdown is similar to investing in value stocks or buying something on sale. Buying a stock at $25 instead of $50 obviously saves you $25 per share, but you can also make the $25 in profit should the stock return to its original price. As the old saying goes: If you liked it at $50, you'll LOVE it at $25. Similarly, starting a a system that is in the midst of a 30% drawdown not only means an investor saved losses of 30% by not getting involved with the system until those losses were realized, but also means that investor will make 30% in profits when and if the system returns to its original level.
If we believe in a trading system and believe that technical based systems should have results that "revert to the mean", then we will see trading systems results prove the old axioms: "what goes up must come down", and the less often heard "what goes down must come up".
Why do systems "Revert to the Mean"
Unlike stocks or funds, a trading system investment won't benefit from increased money flow. One problem with investing in stocks that have been going down is that stocks reflect what people are willing to pay for it. A trading system operates independently of how much money is coming into or out of the investment, crunching its technical formulas on the current market environment, whether that market be soybeans or S&Ps. The problem(or benefit) inherent in a mechanical system is that it is designed to do well in a particular type of environment, meaning it is not likely to do well outside of that environment.
The same thing that makes investing in "hot" systems dangerous make investing in beaten down systems advantageous, and that is the fact that environments change. Governments, economic cycles, weather, science, and people change the way the world goes round every day, causing the specific environment your system likes to either be in or out of phase. A change in environment could be just what the doctor ordered for a system in a drawdown. The tough part, as always, is knowing just when that change occurred.
The benefit of investing in a system at or near its historical Max DD is that the system has most likely just lived through one of the least desirable market environments it could imagine. With conditions failing to line up for the system - it just didn't do well. Will things continue to go against that system? Will market conditions NOT change? Don't bet on it. To the contrary, bet on them changing.
A limited downside:
The attractiveness of buying a stock after its been beaten down or investing in a trading system at or near its maximum drawdown is not just the potentially unlimited upside, but is also about the limited downside (or at least less of a downside than before). This is especially important for those investors who are saying, "But what if it never makes any more money? What if the system keeps going down, adding to its drawdown?"
I'm glad you asked. This limited downside is especially pertinent in a trading system investment, because a trading system's backtested hypothetical results give investors the ability to preset a "line in the sand". The "line in the sand" is nothing more than a worst case scenario level at which point an investor should consider the system broken and stop trading it.
Attain uses several measures for calculating "lines in the sand" for each system - but the math usually works out in such a way that the line in the sand is usually 1.5 times the tested, prerelease maximum drawdown of the system in question.
So, for an investor thinking of investing in the Compass system while that system was in a DD that lasted from July of 2003 until the end of last month, the risk of doing so was simply the difference between the current DD level and the line in the sand level. When the system eclipsed its past historical DD of 52%, the risk in investing in Compass was just 26% (78% - 52%). The potential gain over the following 12 months was the sum of the 52% DD and the system's average annual return of nearly 40%, while the risk was just 26%. That's the potential for 92% gains and a risk of just 26%, or nearly four times as much profit potential than risk. This isn't Vegas, but you won't get much better odds in any investment
But how would one go about investing in a drawdown? And what kind of testing has been done on this strategy? For starters, how do you know when the drawdown has hit its valley? Knowing when the drawdown has it its low point is impossible to know until after the drawdown is over, so we decided to run tests on a drawdown strategy assuming an investor started a system on the trade after it eclipsed it past historical DD. So, in the example above, we would test how Compass would have done for an investor starting the day after the system eclipsed its past historical DD number of 52%.
To see how this strategy of starting a system once its past historical DD was eclipsed had performed in the past on actual systems; we ran tests on the actual and hypothetical results of over 57 systems to find the following results:
Without using a "line in the sand", about 61% of systems saw positive returns between the time they eclipsed their past historical Drawdowns and today.
Using a "Line in the Sand", only 38% of systems saw positive returns between the time they eclipsed their past historical Drawdowns and today. (The rest were unable to continue trading after hitting their stop trade levels).
Using a "Line in the Sand", the average positive return was 58.54%, while the average loss was just -10.08%, showing the limited downside profile expected from limiting losses with a "line in the sand".
Worried that our data may be at risk of some survivorship bias, as many losing systems are quit long before they reach their past historical Max DD and many other systems are not traded long enough to get back to new equity highs, we took things a step further by looking at a few systems which have eclipsed their past historical Drawdowns in actual trading.
The first group of systems shows four systems which eclipsed their past historical Max DDs, but which never hit their stop trade levels. The average system on this list returned 25% after going past the old Max DD, and only had a drawdown since that point of 18%.
The second group of systems are all systems which are no longer traded at Attain because they hit their stop trade levels (represented here by the 1.5 times max DD level). It is important to see how this strategy failed on these systems. Investing in the drawdown proved unprofitable for these systems - but the important thing is losses were contained to a mere 11%
At the end of the day, the data appears to support investing in systems in Drawdowns, as the possible returns are roughly double the possible losses. The following systems have recently eclipsed their past historical Max Drawdowns, making them prime candidates for such a strategy:
Eclipse eRL
R-Mesa 5
Clipper eRL
- Walter Gallwas
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 |
Feature | Week In Review | Chart of the Week |
The month of August is typically very slow from a trading perspective, as traders and investors leave the investment world behind for summer vacations, a beach and a good book. However, there was enough market activity in the the last two weeks of August to more than made up for the market slow down seen at the beginning of the month. Unfortunately, trading picked up because of Hurricane Katrina and the tremendous ripple effects its damage caused throughout the US economy and the world.
Consumers immediately felt the effects of the storm at the gas pump, as prices rose to over $3.00 per gallon nationally. Traders also watched prices soar with Crude Oil futures climbing +11.80% in August, Unleaded Gas futures jumping +38.61% and Natural Gas futures rising an unbelievable 44.85% for the month!
As expected the US Stock market turned around and headed south as energy prices climbed. SP futures finished August down -1.25%, NASDAQ futures were down -1.71%, Russell 2000 futures moved -2.19% lower, and Midcap 400 futures dropped -1.45%. Elsewhere, Japanese stocks looked strong all month as the Nikkei 225 futures gained +5.35%
US Bonds rallied as the stock market sold off. Rumors of the Fed temporarily pausing its quarterly rate raise or even possibly cutting rates caused bonds to rally. 30 year bonds gained +2.66% during August, while 10 year notes were up +1.79%. Across the Atlantic European debt also continues to climb as the benchmark Eurobund gained 1.52%.
The storm and its aftermath also caused foreign currencies to rally, as the prospect of massive US government aid for the area was translated as more US debt and therefore a weaker US dollar. For the month, Eurocurrency gained +1.59%, the Swiss Franc was up +2.45%, and the Japanese Yen moved 1.28% higher against the US dollar. Meanwhile the US Dollar Index fell -1.89%.
Finally, the grain markets saw wheat, corn, and soybeans all traded lower in August as the storm threatened fields and distribution - with wheat falling -7.64%, while corn and soybeans both dropped -1.28%. In the tropical markets cotton futures fell -5.22%, while sugar rallied 2.95% higher.
*Day Trading**
August was a brutal month for systematic traders in general, but a handful of day-trading systems were able to capitalize on the few sporadic days of trending conditions in the markets. The last week of the month helped a few systems rally into profitability and other systems merely recover from the poor start to the month.
The few systems that came out on top for the month were able to filter out the market “noiseâ€-choppiness that often occurs in the slow trade of the summer months. Perhaps the best example for the month was Compass, which profited $3,730.68 and briefly eclipsed new equity highs last week. The system picked its entries wisely, trading an average of just three times per week for the month. R-Mesa eRL is also gaining attention, after profiting $1,869.20 per contract. R-Mesa SP had an impressive run as well, gaining $1,540.25 and ending the month on a positive note with a huge winner on the last day of August. Rounding out the other profitable systems was Impetus eRL, which made a more modest $502.10 per contract but continues its slow climb out of drawdown.
Another group of systems were unprofitable for the month but definitely hung in their for the most part, limiting their losses. The Electric Daybreaker portfolio was unprofitable by -$270 after being dragged down by the eRL and eMD markets.
Spectrum eRL and SP both came out of hibernation but lost -$340 and -$450 a piece. The systems trade very infrequently but have proven to be valuable systems to add to a portfolio for diversification. Nautilus ES traded just over once a week for the month but lost -$580 per emini contract. The system works off similar underlying logic as Compass but has several unique differences. The system can and will enter the market in the early part of the session unlike Compass and also will jump right in line with a trending market without waiting for a retracement. Helix ES performed well on choppy days but suffered some big losses on days with large moves. The system traded thirty seven times in August yielding a loss of -$622.50 per contract. RC Success ES suffered its third consecutive losing month losing -$681.25 per contract.
The final group of systems suffered larger losses in August. Clipper eRL traded five more times than sister-system Compass on its way to losses of -$1,093.50 for the month. We always preach the importance of diversification, and customers trading both Clipper and Compass saw how different the two systems can perform despite similar underlying logic. BWT Zones eRL got caught up in several unprofitable reversal trades that led to losses totaling -$1,144.10 for the month. Daybreaker SP suffered its third consecutive losing month with losses of -$2,211, while BWT Zones SP cooled off after two monster months in June and July with losses of -$3,250. Finally, BWT Rock N Russ was unprofitable by -$3,981.60 per money management unit using the position manager.
**Swing Trading**
August 2005 may go down as one of the worst months in history for the broad spectrum of swing trading systems. Going down the list of systems the only programs to turn a profit were in the bonds. Mesa Bonds topped off the month with strong gains in open and closed trade profits of +$2,853.12. Mesa Notes was not far behind earning +$2,731.25 for the month, followed by the Jaws Narrowneck Portfolio earning +$268.75.
By far the hardest hit system was Eclipse eRL which was thrown into a whirlwind of low volatility, institutional eRL trading errors, and low volume. The system utilizes various forms of technical analysis in electing its trades, including heavier weighing toward up and down volume– During a 6 day stretch from Aug 10-16th the market experienced 6 different changes in volume direction adding to the systems open and closed trade losses of -$7,924.40 per contract.
Beyond Eclipse Axiom and Tzar also struggled. In open and closed trade losses the 4 market (ES, NQ, eRL, eMD) Axiom Index portfolio lost -$3,841.20 and the Tzar 4 market portfolio lost -$7,137.50. On a market by market basis Axiom Index finished as follows; ES -$2,370, eMD -$1068, NQ -$760, and eRL +$356.80. Tzar’s individual markets finished -$2,840 in the eRL, -$1,947.50 in the ES, -$1,300 in the NQ, and -$1,050 in the eMD.
Rounding out the battle wounds of August were the Trade Maid and Roker Capital suites of systems. Trade Maid's Bounce eMD lost -$1,070 on 4 trades and Bounce eRL lost -$370 on 2 trades. Roker’s Apollo ES ended down -$1,535 on 7 trades and Athena eRL lost -$1,050 on 3 trades.
**Long Term**
It has been a tough year for long term trend followers and the month of August may end up being the exclamation point on one of the worst years in recent memory for long term, trend following system traders. Overall, choppy market conditions dominated across nearly all of the commodity markets, while trend reversals in the bonds and currencies hurt systems with short positions.
One system that continues to thrive despite the volatility (or lack thereof) is Axiom LT. Axiom LT, which has been traded at Attain since January, has had the best year out of all trend following systems. Axiom’s ability to jump on trends early has allowed it to rack up gains in foreign currencies, energies, and grains. The system also has broad exposure to a variety of foreign markets like London Sugar, London Coffee, the LME metals, and foreign bonds, as a group these markets have provided smoother trends than their US counterparts and Axiom investors have benefited.
Key open positions for Axiom LT include a short Yen trade for open trade profits of +$3968.75 per contract, a short cotton trade for profits of +$575.00 per contract, a long London Sugar trade for open trade profits of +$1965.00 per contract, and a sort corn trade for open profits of +$637.50 per contract, and last but not least a long crude oil trade with open profits of +$6020.00 per contract. It hasn’t been all roses for Axiom however. Every system has losing trades, however; and closed out trades from August for Axiom include a loss of -$3018.75 per contract in the bonds, -$1685.00 per contract in the 10 year notes, and -$1025.00 per contract in bean oil.
Brix is another long term system able to find success with open trade profits of +$3800.00 per contract in the Yen,+$1660.00 per contract in the Eurobund, and +$1575.00 per contract in the Swiss Franc. But - like Axiom LT, Brix found the US Bond market tough to trade losing -$-1737.50 per contract on a short 30 year position.
Aberration Plus has had a rough year and August only added to its troubles with several losing trades in the bonds. Stopped out trades include a short trade in the five year notes for a loss of -$1331.25 per contract in the five year notes, -$2378.00 per contract in the 10 year notes, -$725.00 per contract in the 2 year notes. The system did close out a few winning trades including a gain of +$2600.00 in the Peso and +$1370.00 per contract in the Dollar Index.
Andromeda also struggled in the bonds losing -$1315.00 per contract in the five year notes, -1587.50 per contract in the two year notes, and -$375.00 per contract in Eurodollars. Andromeda also closed out a profitable position in the Dollar Index with the system making +$1325.00 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.