HOUSE_OVERSIGHT_026012.jpg

While this particular example above illustrates good results by the end of the day (+$7,848 and $1,600 maximum drawdown from starting balance), there are certainly markets situations that require large margins and can not be expected beforehand. These are usually terrorist attacks, sudden nature disasters, or even FOX News releasing FBI renewal investigation over Clinton and following speculation on Trump’s lead prior U.S. elections.

The Brexit day example is certainly a good case scenario for my algorithm. It provides algorithm with a large volatility and many profitable opportunities. However, there are other large one-side price movements in financial markets that can destroy not only my model, but many other trading strategies if wrong side position is held. In the last month financial markets experienced pound flash-crash in the middle of a random night. GBP/USD dropped for about 10 cents in 2 minutes and jumped back. Many got rich and many got poor for 120 seconds while in their sleep. The only thing that secures investments is responsible trading, which is placing stop losses and not risking more than some percentage per trade.

As a CFA Candidate and a prospective algorithmic hedge-fund manager I cannot say investing even in T-Bills is 100% safe and can only say currency trading involves high risk. The only thing I can assure tell you (in my personal opinion) that $10,000 is a lowest safe responsible sum of money needed to operate with the model where $5,000 is used for margin alone and $5,000 is set to be used as a bad case scenario expected drawdown value or a stop loss.

P.S. Jeffrey, I want to introduce you to some algorithmic trading concepts and terms needed to understand trading algorithms better.

• Algorithmic trading is highly dependent on statistics and averages over a long-term history.
• Drawdown – maximal losing dollar amount for open or closed positions from starting balance.
• Algorithm ceiling – largest amount of money an algorithm can work with (large trades can influence market conditions against an algorithm).
• Percentage of profit and loss trades: 53.99% profit and 46.01% loss trades this summer on a “back-test”
• Average profit trade: $20.74 this summer on a “back-test”
• Average loss trade: $14.13 this summer on a “back-test”
• Number of trades per time period: 45,221

P.P.S. Drawdown value for a condition where price goes down, my algorithm buys a long contract every minute and does not hold short contracts can be estimated in the Excel file attached to the email (Estimation of losing P&L).

2

HOUSE_OVERSIGHT_026012