Market-neutral investment strategy

Information presented here is for educational purposes only and does not intend to make an offer or solicitation for the sale or purchase of any specific securities, investments, or investment strategies. Investments involve risk and there are no guarantees of any kind. Be sure to first consult with a qualified financial adviser and/or tax professional before implementing any strategy discussed here in. Past performance is not indicative of future performance.

Making an investment strategy that is overall uncorrelated to the stock market

Our vision

  • Customised solutions
    • Less number of uids ( independent portfolios ), so that we can manage them better.
    • Simple tool to technical sales team to engineer solutions.

Absolute Return

Total Return

Customised solutions

RoadMap

 

  • Strategy Development infrastructure for daily data
     
  • Long only mandate portfolios
  • Score functions aligned with alpha strategies based on deep learning
  • Initial CIO layer - based on risk and correlation
     
  • Score transformation
  • Move to single position sizing 
  • CIO layer version 2 
  • More efforts on deep learning
  • Long Short ETF portfolio

2015

2016

2017

Current Workflow

Investment Universe

Low expense, high volume, liquid ETFs

Score

score assigned to each product denotes goodness of investing in it

Position sizing logic

Translates scores to positions with the aim to achieve good returns while managing the risk and adhering to constraints

 Risk Management

Monitors and controls the portfolio drawdown

Systematic CIO layer

Allocates to strategies with better chances of outperforming in future while adhering to portfolio mandate

Strategy 1

Strategy 2

Execution

Execution plays a key role while managing a high turnover portfolio. Also, a good execution system can add alpha on top of portfolio returns.

 

Roadmap

  • Infrastructure

  • Data Processing
  • Passive Fill Strategies : Mars & Saturn
  • Passive TWAP: splitting Orders across day
  • Model Aggregator

Ongoing Work

  • Deep Learning

    • FNN Prediction strategy 

    • Transform Intraday data to augment data daily
       

  • Next Version of Systematic CIO layer
    • Regime prediction
    • Diversification across risk premia
       
  • Reinforcement Learning based execution
     
  • Single Name Stocks Portfolio

End of Part1

Portfolio Details

  • Detailed results can be seen here

Assumptions

  • Annual Maintenance cost of 3 bps on the dollar amount traded is incorporated into results
    • This includes commissions, borrowing cost and slippage
       
    • Example : For an initial portfolio of 400k with turnover = 5000%                                    annual maintenance cost = 12K ( 400000*50*2*0.0003 )
       
    • For IB, we expect trading commissions of 5K-6K per year.
      3.5K = ( 40 * 250 * 0.35 ) = ( num_products * num_trading_days * min_cost_per_order )

      For larger portfolios, this cost would be low in terms of percentage.
      But it puts a lower bound on the AUM in an account ( if it is a SMA )

       
  • Execution at the open price
     
  • Portfolio Value >= 400K

Long Short Portfolio Mandate

Param Constraint Is Customizable
Leverage <=4 Yes
| Long - Short | <=1 Yes
TargetVol 10% Difficult to increase
Rolling Beta to SPX -0.2 to 0.2 Yes
Turnover < 6000% Yes

Long Short

By qplum

Long Short

  • 1,300