.png){kind=small}
The trading systems showcased on this website are now defunct and non-operational w.e.f. May, 2022. This is due to restrictions related to conflicts-of-interest in my current professional role.
Proprietary trading
systems
Multi-strat quant trading systems for the Indian derivatives market
Market profile, price action, market microstructure and option greeks
Backtesting, simulations, walk forward optimization and quantitative filtering
Reliable estimates of the expectancy >> Optimal position sizing
Mechanics
Process
Edge
.png)
Cauchy distribution is special case of the student-t with 𝑣 = 1
pdf of a Cauchy random variable is as follows:
STATISTICS
What is the Cauchy Distribution?
Cauchy Distribution is a fat tailed continuous probability distribution where extreme values dominate. It is a special case of the Student's t-distribution with 𝑣 = 1.
Cauchy has no finite variance and no moment generating function. If one is drawing samples from a Cauchy population and computes the sample mean and sample standard deviation, they should never see 1/√n behavior.
Neither the Law-of-large-numbers, nor the Central-Limit-Theorem applies to the Cauchy Distribution - even though the Mode is defined as 𝜇.
TRADING SYSTEM
Overview
Nifty50 day ranges are not actually expected to follow a Cauchy distribution. The idea is that we build a dynamic system that identifies day structures (using Market Profile) and opens exposure to either short vol or long gamma strategies based on its initial identification. Empirically, it has been observed that large range days are pretty unbounded, relentless and pathological, like the Cauchy Distribution.
Going through price data in real time (using real-world probabilities), we don’t know what type of distribution we are working with. What we do know after we have collected a reasonable sample size (described in time terms) is what type of distribution it is likely not. Once a normal profile in Market Profile terms has been determined to be improbable, the system works with the assumption of the most extreme distribution curve that there is. Risk, position-sizing and profit-taking parameters in the system are then appropriately adjusted.
The system is primarily run on Nifty50 Options and relies on a volatility based position-sizing and risk management framework.
system development
Process
Backtesting
HISTORICAL PRICES
Conduct preliminary backtests without any optimization (risk management through price stops). Followed by non-parametric simulations.
Quantitative
filtering
Introduce Market Profile based probabilities by using day-type and time-based filters.
CURRENT PRICES
MARKET PROFILE STATS
Walk Forward
Optimization
12 months of disciplined walk forward optimization including realistic entries/exits in a live market. Other actionable stats are also collected during this process.
DISCRETION OUT
Rules & guidelines
Optimisation of the walk-forward parameters >> framework and rules for each of the setups within the system.
Performance Measurement
Clinical assessment of the sources of returns (return attribution) is essential for managing risk for an Options book. Since regular non-convex measures of performance cannot be used here, below are some of the performance variables that I track closely on a monthly basis.
Used to identify if the correct profit taking parameters are used for each of the setups.
Compare real vs. theoretical
Used to identify if the correct stops are used for each of the setups.
How much money is being left on the table?
3 month rolling performance of profit taking methods
When holding any longer has diminishing marginal returns
Rupee gained per rupee risked is the actual risk multiple of the trade. This also tracks execution/discipline in stop taking and benefits of the discretionary exits.
Best to worst comparison of the trade while the trade is on
Basic P&L and equity curves for each month
Advanced actionable insights from each month’s journal
RESOURCES
-
MARKET PROFILEJames Dalton – Mind Over Market, Markets in Profile Peter Steidelberg – CBOT Handbook
-
PRICE ACTIONAl Brooks – Trading price action Reversals Al Brooks – Trading price action Trends Al Brooks – Trading price action Trading Ranges Trader Vic – Methods of a Wall Street Master Lance Beggs – YTC Price Action Trader
-
QUANTITATIVE METHODSLarry Williams – Long term secrets to Short term trading Kevin Davey – Building Winning Algorithmic Trading Systems Shreve 1 - Stochastic Calculus for Finance (Discrete) Shreve 2 - Stochastic Calculus for Finance (Continuous) Paul Wilmott - Introduces Quantitative Finance Measure Theory and Stochastic Calculus Playlist Econometrics Playlist Pricing and Computational Finance Playlist
-
OPTIONSSheldon Natenberg – Option Volatility and Pricing Nassim Taleb – Dynamic Hedging Colin Bennett – Trading Volatility: Trading Volatility, Correlation, Term Structure and Skew
-
POSITION SIZING AND RISK MANAGEMENTVan K Tharp – Definitive Guide to Position Sizing Aaron Brown – Red Blooded Risk
Raghav Duseja
ABOUT ME
BBA LLB | MScFE | CFA | FRM
Previously engaged in developing and running these proprietary trading systems for NSE/MCX derivatives on my own account.
linspace.me
Macro research blog which covers the global monetary system, energy complex and Indian derivatives
BLOG
I am not a SEBI registered research analyst or investment advisor. All the information provided here are for educational/informational purposes only. CauchyDistribution is strictly proprietary and wholly self funded and does not accept or manage any public funds or portfolios.
© 2022 by CauchyDistribution.
The trading systems showcased on this website are now defunct and non-operational w.e.f. May, 2022. This is due to restrictions related to conflicts-of-interest in my current professional role.
RAGHAV DUSEJA