Sip Optimizer(Kinda)

This is a backtesting framwork(again, kinda) to understand if doing SIP with any kind of strategy that people talk about are worth looking into or not.

The Current implementation takes data in CSV:

"Date","Open","High","Low","Close"
"19 Aug 2022","17966.55","17992.20","17710.75","17758.45"

and converts it to this format:

{                     
  Date: '19 Aug 2022',
  Open: '17966.55',   
  High: '17992.20',   
  Low: '17710.75',    
  Close: '17758.45',  
  momentDate: Moment<2022-08-19T00:00:00+05:30>,   
  UnitPrice: 17758.45 
}

Observations regarding Data:

1. UnitPrice is considered same as Nifty Closing price of that day.
   • Can change it to ur favorite Fund's Daily nav
2. momentDate is a internal date feild that converts the Date to a moment.js format
3. Currently Nifty50 10yrs data is loaded into the script

The Trading Service:

 /**
   *
   * @param {Number} startDate 1-31 start sip on this date
   * @param {Number} sipAmount
   * @param {Boolean} enableAlgo should enabled the formulae or not
   * @param {typeof Formulae.formulaToBuyOn} formulaToBuyOn callback function
   * @param {typeof Formulae.multiplierFormula} multiplier callback function
   */
  const tradingService = new TradingService(
      START_DATE,
      SIP_AMOUNT,
      true,
      // THis is the expression that needs to hit in order to place buy order
      Formulae.formulaToBuyOn,
      // This is the expression that multiplies the SIP amount according to the last multiplier and how many times the formula hit in a month
      Formulae.multiplierFormula,
    );

The parameters:

1. Start date which day of the month you want to start the SIP
2. Self explanatory SIP amount
3. if you want the algorithm/strategy you want to be enabled or not. Works if you want to measure the base xirr
4. a Callback for the formula/Strategy/algorthm you want to create. It is an expression so returning truthy value means there will be a buy call
5. a Callback for the multiplier of the SIP amount. For example if you want to invest double the amount if your formula hits then you can set it here.

Default Formulae/Strategy/Algorithm

1. The algorithm currently baked into this is: if in a month the nifty index falls 500 points then BUY.

/**
 * The expression if true a buy call will be placed
 * @param {Number} startingNiftyThatMonth at the price the first sip was bought
 * @param {NiftyData} niftyData
 * @returns
 */
static formulaToBuyOn(startingNiftyThatMonth, niftyData) {
  const actualThreshold = THRESHOLD * ((100 - PERCENTAGE_ACCEPTABLE_DELTA) / 100);
  return startingNiftyThatMonth - Number(niftyData.Close) >= actualThreshold;
}

2. The multiplier algorithm: First 500 buy SIP amount, if it falls another 500 invest 2x SIP amount, another 500 4x SIP amount. so, X,2X,4X and so on.

/**
 * if want to increase sip amount if the formula hits
 * @param {Number} lastMultiplier
 * @param {Number} ithForumulaHit the formula hits for that month. Zero based. So the first Hit will be zero.
 * @returns
 */
static multiplierFormula(lastMultiplier, ithForumulaHit, niftyData) {
  if (!ithForumulaHit) return 1;
  return lastMultiplier * 2;
}

Running:

1. Clone the repo
2. npm install.
3. npm start

   % npm start
   
   > sip-optimizer@1.0.0 start
   > nodemon -x "npx babel-node src/index.js"
   
   [nodemon] 2.0.19
   [nodemon] to restart at any time, enter `rs`
   [nodemon] watching path(s): *.*
   [nodemon] watching extensions: js,mjs,json
   [nodemon] starting `npx babel-node src/index.js`
   ----------------------
   AlgoEnabled: true
   Threshold: 500
   Current Value:  1709702.3957544165
   Invested Value:  885000
   Profit:  824702.3957544165
   XIRR:  14.58731463444985
   ----------------------
   [nodemon] clean exit - waiting for changes before restart

Note: Can probably use it as node module. Didn't test.

Further Development and Future plans for this repo

For further development, feel free to fork the repo.

1. Current plans for this repo are creating a chart with different threshold amounts and different strategies and put the chart in the Readme itself.
2. Instead of manually downloading the data, injesting an API(preferably free) to get the data instead.

Disclaimer: Its not like an investment tool, its a pet project for backtesting some interesting claims people make. Given that there are other tools for probably 1000x better if you want to seriously do algo-trading then I suggest you use those.