Methodology

Random Walk Capital is a fictional long-only hedge fund seeded with $50,000 capital. Every working day (Monday to Friday) a random stock pick is made by an algorithm that selects from a selection of over 20,000 equities actively traded on the New York Stock Exchange, NASDAQ, AMEX, London Stock Exchange, Tokyo Stock Exchange, Toronto Stock Exchange, Frankfurt Stock Exchange and Australian Stock Exchange. The algorithm further determines position size and holding period on a completely random basis.

No AI or machine learning is used in the decision making process. The fund is purely an experiment to demonstrate the Random Walk Hypothesis in practice.

This project was created by Charlie (https://charlie.onl) as an educational tool & hobby project.

The fund exists to demonstrate the “Random Walk Hypothesis,” a financial theory which says stock prices move randomly and future price changes cannot be predicted. By making every part of each trade random, apart from how often we place BUY trades, we can see how a truly random long-only hedge fund might perform in the real world with a realistic amount of capital for a retail investor.

Trades are made daily by an automated script which tries to represent a truly random trading decision being placed. Position sizing is calculated via two algorithms. One being a Cash Allocation Algorithm that determines a random amount of cash to allocate to a trade (between a minimum and maximum size) and the second method being a slightly more sophisticated one that considers a pre-determined risk level and acceptable stop-loss. Once a stock & position size have been decided another algorithm calculates the minimum number of days the stock must be held before it can be sold and the maximum number of days that the stock can be held for before it must be sold. Once these factors have been decided the fund will simulate a trade based upon the latest market price available on Google Finance and will add it to the list of “Holdings” held by the fund. The amount allocated to the trade will be deducted from the fund’s fictional cash pot.

There are numerous ways that the fund will dispose of positions. Firstly, whenever a trade is placed the system will calculate a minimum and maximum number of holding days. The minimum number represents the shortest time frame that the stock must be held in inventory before it can be sold. The maximum number represents the number of days that the stock can be held before it must be sold. The automated script responsible for handling BUY trades will first run a check for positions that must be sold that day and will offload those first. Another method of determining when to SELL a holding is by having a maximum number of holdings that the fund can hold at a single time. If the number of holdings in the fund exceeds that number when the trading script is running then there is a random chance that a stock will be sold. This will either be NO sale (unless the fund is holding the maximum number of holdings plus an additional 5, at which point sale becomes mandatory), or the script will randomly decide to sell either the highest value holding in the fund or the longest held position in the fund.

The system uses Google Finance price data when making buy/sell decisions. It does not consider the spread (bid/offer) prices. These prices may not be entirely accurate but should represent a close enough picture for us to be able to determine long-term performance of a truly random fund. There may be need for manual intervention if something in the system suddenly breaks, such as our APIs responsible for fetching stock prices or currency rates. All manual interventions will be logged and will try to be kept to a minimum. These decisions may not reflect accurate real world returns due to the limited datasets we have access to. Trading decisions made by the fund DO NOT count as financial advice. Random Walk Capital is a fictional fund made for educational purposes to experiment with financial theory in a practical way.

NOTICE: The fund does not consider transaction costs, taxes, or fees when making trading decisions. In the real world, these factors can significantly impact net returns.