short term momentum pdf

12
Short-Term Momentum A Ten Market Combination Clark Collins 2015

Upload: clark-collins

Post on 15-Aug-2015

47 views

Category:

Documents


1 download

TRANSCRIPT

Page 1: Short Term Momentum PDF

Short-Term MomentumA Ten Market Combination

Clark Collins

2015

Page 2: Short Term Momentum PDF

Momentum Trading can be defined in several ways as one can learn through Investopedia ™ but for the purpose of this shortbrief, I have used the word momentum in place of what are actually “trend following” concepts applied over a short -termtime horizon. Momentum trading follows very similar concepts to trend following but is normally utilized in “day trading” andpositions are rarely held longer than the initial day of trade origination. Utilizing simple trend following rules, I have appliedthem to intraday data, without the mandate to exit before day’s end such as the momentum trader generally does. After atwo-year learning period where a best parameter selection was determined, I then walked forward in a blind test to reviewhow well the strategy performed over historical data. Utilizing ten non-correlated markets and applying some basic portfolioconcepts, I have documented those results for review here.

Trend following is one of the oldest and most popular quantitative applications in the alternative financial space. Virtually allasset managers who utilize quantitative investing approaches today began their careers by cutting their teeth on trendfollowing concepts taught in graduate or doctoral school while reviewing time series analysis. The reason that this genre is sowell-covered is because of its long and profitable history with past and present practitioners of the trade.

The simple facts are that trends exist, even with incredible price noise/volatility present in the markets today. Good trendfollowing rules continue to work despite the inherent volatility the trend follower must endure. The marketplace still exhibitslong-term cycles, whether the result of geopolitical events or fundamental supply and demand changes. When those largevaluation shifts are sustained for reasonable periods of time to new equilibrium price levels, profits for the trend follower canbe achieved.

As long as there is uncertainty in the global marketplace, we will see these cycles/price patterns evolve with randomness butnonetheless continue along the many diversified sectors of the world economy.

Page 3: Short Term Momentum PDF

The Ten Market Portfolio

Deciding on which futures markets to use for this study, my main concern was NOT to only select those markets which hadshown a propensity for trending behavior. To avoid that pitfall, I chose a decent handful across multiple sectors with verylittle to no relationship or dependency on any one econometric factor.

Performance bias with regards to quantitative system engineering is a real pitfall and expectations of future performance cannever be predicated on assumptions known of past history. An example of this would be to present a long-only strategytested over the past five years, which trades the U.S. domestic equity market. Who could not or has not made money goinglong or buying the U.S. equity market these past five years? To posit an assumption that one could expect similar results forthe next five years based on a study over what we have all experienced in the domestic equity market these previous fiveyears would be a bias of incredible proportion, and an investment wrought with trouble for the impending investor.

The ten markets selected, represent seven different sectors, where each of the ten has an equal stake in their impact in thistheoretical portfolio. These sectors are currencies, interest rates, domestic equity, softs, energy, grains and precious metalsmarket. Many other well-known individual markets were not included, but when applied to this current basket, the additiondoes help the overall risk-adjusted performance as one might expect. A ten market portfolio seemed like a manageablenumber, and I believe the results are interesting to begin a dialogue.

Page 4: Short Term Momentum PDF

The Markets The Costs

Australian Dollar - Currency Trades assumed a cost of $15 per round turn

Crude Oil - Energy Spread costs assumed the same $15 per round turn

Cocoa - Soft Management fees per month on overall portfolio at (1%/12)

Euro Currency - Currency Incentive fees per month at 20%, on New Profits Only

Emini - S&P 500 - Equity Slippage costs not applied as execution on hourly time

Copper - Metal No Price Stops, only time exit when trade levels are triggered

Soybeans - Grain

Silver - Metal

Sugar - Soft

U.S. 30 Year Bond - Interest Rate

Page 5: Short Term Momentum PDF
Page 6: Short Term Momentum PDF
Page 7: Short Term Momentum PDF
Page 8: Short Term Momentum PDF

Trade and Portfolio Processes

Trend following and momentum trading approaches are intuitive. They require a measured shift in price direction for thestrategy to be engaged and establish positions along the relative price trend. Another way of explaining this is simply sayingthat price action must show significant strength to BUY or significant weakness to SELL and ultimately originate a trade. Thisstrategy does not buy bottoms or sell tops. It is not predictive. Like all trend following and momentum strategies, once thetrade is entered, the strategy is profitable if the trend continues. If the trend breaks down and consolidates, the trademanagement process either holds its position in the noisy environment or exits the trade at a loss.

Trend following and momentum concepts must follow the defined price direction of the time series window being analyzed.To be more specific, if the trend follower or momentum trader is analyzing daily or weekly data and attempts to capturelarger macro shifts in price in an unmistakable upwardly moving market, the momentum and trend should dictate a longposition for the manager or trader. A trend follower could and frequently does manage risk along that trending cycle but theoverall trade exposure in that specific market must be with the trend.

Utilizing intraday data, I have established a simple two parameter algorithm to monitor short-term trends. Because of thisshorter time series, compared to long term trend followers, I can establish short positions or long positions which can looklike they are against the longer term trend but when reviewed in its relevant time scale, what I have followed is unmistakablytrend following and momentum based.

The next page illustrates some trading examples visually and against the prevailing longer term price trend in some instances.

Page 9: Short Term Momentum PDF
Page 10: Short Term Momentum PDF

MOMENTUM TRADING MONTHLY NET % RETURNS(Fit Data Highlighted)

Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec YTD2008 -0.40% -0.08% 3.11% 1.92% 3.77% 3.50% 1.95% 2.58% 2.85% 20.80%2009 -2.89% 1.66% 1.20% 3.24% 3.05% -0.09% 4.45% -0.50% 2.29% 1.38% 3.59% 3.19% 22.34%2010 1.04% 0.38% -1.08% 1.75% 2.47% 1.43% 2.17% 1.46% 2.06% -1.33% 3.04% 2.31% 16.77%2011 0.57% 1.64% 0.63% 1.79% 1.16% 1.89% -0.07% 3.91% 4.71% 0.92% -1.62% 0.55% 17.15%2012 0.46% -1.07% 1.23% -0.87% 3.47% -1.34% 1.10% 3.50% 3.10% -1.30% -0.73% 2.92% 10.76%2013 -0.29% -4.00% -1.76% 2.30% 0.65% 2.47% 0.09% 0.38% 0.95% -0.99% 0.69% -1.01% -0.70%

Correlation of Momentum Trading to S&P 500 -0.074Correlation of Momentum Trading to Barclays BTOP Index -0.014

Page 11: Short Term Momentum PDF
Page 12: Short Term Momentum PDF

ConclusionSo much about this short brief can be elaborated on and the higher moments of statistical analysis applied.For this brief paper however, I only wanted to show a short and powerful process which can be implementedquickly and understood easily. I have purposely left out trade statistics which also help to validate theprocess as momentum or trading with a price trend. Those metrics are very similar to what you would findwith most managers utilizing this style of trading.

Value at Risk (VAR), Monte Carlo simulations and other metrics can easily be applied to help validate andaccentuate this simple but powerful trading process. I have also left out capacity studies as well as positiveand negative slippage reports where applicable. Needless to say, assets under management (AUM) capacityin these ten markets with this relative style of trading could be maintained well over 300 million USD withoutimpacting market slippage. Adding additional markets will also address that concern if it ever arises.

Please feel free to contact me at your convenience about any questions or comments about this brief. Iwould be very interested in having an additional conversation about implementing this as a potential product.

Best Regards,

Clark [email protected]