//
Strategies

All strategies table 9Above is the performance comparison of various systematic trading algorithms / strategies which are described below. The algorithmic models are based on different criterias giving buy or sell orders. The comparison is for last 15 years, from 1999 until 2014. Companies invested have minimum market cap of USD 100 million and OTCs, ADRs, foreign / non US stocks and less liquid companies are ignored / excluded. Returns are net of commission (0.1% for every trade) and slippage (0.3% for every trade).

 

Strategy 1: Multiple Factors

Description: This systematic trading algorithm is based on multiples factors such as low valuation, low short ratio, momentum, asset light, high return on asset, high return on debt and other variables. The algorithmic model rebalances itself weekly and ranks stocks based on the specified criteria’s. After ranking it gives buy, sell or hold orders. 

Criteria for selection: Companies invested have a minimum market cap of at least USD 100 million. The portfolio maintains 10 stocks at any given point. The model ignores / excludes OTCs, ADRs, any foreign / non US stocks and less liquid companies.

Buy and sell: If the rank of an existing stock in the portfolio falls below the specified threshold, the model gives sell order and buys stocks which are currently ranked higher. Once signals are generated, the stocks are bought and sold at next day’s opening price.

Hold: If an existing stock in the portfolio maintains its rank above the specified threshold the model holds the stock.

Commission and slippage: Model includes 0.1% commission and 0.3% slippage for each trade.

Performance: In last 15 years (from 1999-2014) the algorithm/model has significantly outperformed S&P, generating an exceptionally high return of approximately 40%+ cagr (net of commissions and slippage), as compared to 3.3% for S&P during that period while consistently beating the market every year.  Average holding period of a position is approximately 25 to 45 days. Below are performance and other details:

 

Multiple factors - All - Unhedged - table Multiple factors - All - UnhedgedMultiple factors - All - Unhedged - PerformanceMultiple factors - All - Unhedged - Performance diff yearsMultiple factors - All - Unhedged - Statistical details

 

Multiple factors - Russell 2000 - Unhedged - tableMultiple factors - Russell 2000 - UnhedgedMultiple factors - Russell 2000 - Unhedged - PerformanceMultiple factors - Russell 2000 - Unhedged - Performance diff yearsMultiple factors - Russell 2000 - Unhedged - Statistical details

 

Multiple factors - Russell 3000 - Unhedged - tableMultiple factors - Russell 3000 - UnhedgedMultiple factors - Russell 3000 - Unhedged - PerformanceMultiple factors - Russell 3000 - Unhedged - Performance diff yearsMultiple factors - Russell 3000 - Unhedged - Statistical details

 

Hedge: The above algorithmic models were completely unhedged. The algorithm is also tested with specific hedge rules in place. When those rules are met, a hedge is entered with 50% of the portfolio allocated to shorting S&P ETF (ProShares Ultra S&P500 – SSO). A hedge is entered if the current year consensus EPS estimates are trending down or using VIX as a gauge to determine if an aggressive or patient moving average crossover should be used. Below are the results:

 

Multiple factors - All - Hedged - table

Multiple factors - All - HedgedMultiple factors - All - Hedged - PerformanceMultiple factors - All - Hedged - Performance diff yearsMultiple factors - All - Hedged - Statistical details

 

Multiple factors - Russell 2000 - Hedged - tableMultiple factors - Russell 2000 - HedgedMultiple factors - Russell 2000 - Hedged - PerformanceMultiple factors - Russell 2000 - Hedged - Performance diff yearsMultiple factors - Russell 2000 - Hedged - Statistical details

 

Multiple factors - Russell 3000 - Hedged - tableMultiple factors - Russell 3000 - HedgedMultiple factors - Russell 3000 - Unhedged - PerformanceMultiple factors - Russell 3000 - Hedged - Performance diff yearsMultiple factors - Russell 3000 - Hedged - Statistical details

 

 

 

Strategy 2: Value and Future Growth

Description: This strategy is based on low valuation and high return on capital. The model rebalances itself weekly and ranks stocks based on the specified criteria’s. After ranking it gives buy, sell or hold orders. 

Criteria for selection: Companies invested in have a minimum market cap of at least USD 200 million. The portfolio maintains 10 stocks at any given point. The model ignores OTCs, ADRs, any foreign / non US stocks and less liquid companies. 

Buy and sell: If the rank of an existing stock in the portfolio falls below a specified threshold, it gives sell order and buys stocks which are currently higher ranked.

Hold: If an existing stock in the portfolio maintains its rank above the specified threshold, the model holds the stock.

Performance: In last 15 years (from 1999-2014) the model significantly outperformed S&P, generating a cagr of 26.9% (net of commissions and slippage) as compared to 3.28% for S&P during that period. 

Commission and slippage: Model includes 0.1% commission and 0.3% slippage for each trade.

Value and Future growth - All - Unhedged - table

Value and Future growth - All - Unhedged

Value and Future growth - All - Unhedged - Performance SMALLValue and Future growth - All - Unhedged - SMALL Performance diff yearsValue and Future growth - All - Unhedged - small Statistical details

Value and Future growth - All - Hedged - tableValue and Future growth - All - HedgedValue and Future growth - All - Hedged - small PerformanceValue and Future growth - All - Hedged - small Performance diff yearsValue and Future growth - All - Hedged - small Statistical details

 

Discussion

No comments yet.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

Follow

Get every new post delivered to your Inbox.

%d bloggers like this: