Chart shows the progress of the DataSwarm Trading Fund (DTF) from start in December 2019, tracking market performance from portfolio buildup into the trading stage, and then navigating the market crash and rebuild
Predicting asset prices using the DataSwarm Zeitgeist
Followers of this blog will recall that we have used the DataSwarm Analytics Engine on both qualitative analysis of companies (for example, challenger vs. traditional banks) and prediction of stock prices (for example at IPO). Our experience in the IARPA Geoforecasting Challenge last year (for which we won an award) also proved that our system was very competitive at predicting asset prices.
Since mid 2019 we have been evaluating the system for stock market trading, using the Dataswarm algorithms to predict and track the Zeitgeist – the “bull and bear” movements – in the US stock market.
We spent an initial 2 – 3 months on research and “paper trading” to develop our strategies and test the effectiveness of our DataSwarm analytics on the selection of stocks. We then traded on the US markets from September to November, testing the system analytics and trying strategies ranging from “Buy and Hold” to Day Trading.
Having gained experience and confidence in a live trading environment we progressed to trading a “virtual fund”. For this “DataSwarm Trading Fund” we took advice from professional fund managers about formal rules. Specifically, the key rules we were set were to define:
- Maximum and minimum holding levels of any one stock
- Minimum spread of stocks held
- Limitations in stock types allowed (for example minimum market capitalisation)
- Some limitations in trading strategies
- A defined strategy for risk mitigation
The aim is to ensure that the system’s operations are repeatable and scalable.
Achieving a 10% Return on Investment in “Interesting Times”
The headline performance of our fund is illustrated in the graph at the top of the page, which illustrates the return in capital we have achieved from the 1st December to 18th May, compared to the S&P 500.
As you can see, our system has achieved a c 10 % return over c 5½ months. But this belies a major market crash in the interim that we had to navigate our way through. At the same time the S&P 500 Index fell by about -7 % (and the tech stock heavy NASDAQ by about c -4 %).
Prior to this crash, from the beginning of December to just before the crash in late February the DataSwarm Markets Trading System had delivered a c 13% return on the initial investment – the “Assets Under Management” (AUM) in Fund Management terms – against a rise in the S&P500 of c 7%. (see on chart)
DataSwarm Analytics – Predicting market shifts, rebalancing Portfolios
Our fund fared better than the indices because a system managed fund does not have to follow an index down in a crash. Our systems’ analytic indicators advised a rapid shift into better performing sectors (e.g bio-tech sector and cash). When the markets start to rise rapidly over April, the system drove more portfolio shifting into other stock types.
The DataSwarm System has come through one of the toughest tests you can apply to any stock tading strategy – a major market crash – and is almost back to where we were at the pre-crash top of market.
We are thus quietly confident that the predictive algorithms of the system work to a material level. We have also learned a lot of very useful lessons from this unstable period, which we are using to improve our algorithms even further, and build in new predictive capabilities.
Workflow Tools – Designed for Speed
The other lessons for ongoing system development were a deeper understanding of the tools required to be able to monitor and manage a far larger range of stocks, end to end, in risky conditions. As the portfolio gets larger and more complex – especially in periods of high volatility – speedy workflow and an at-a-glance user interface become increasingly important.
Would you like to use the DataSwarm system?
We have already been testing the system’s ability to give near real time trading alerts for nearly a year, and are now ready to open it up now. Our dashboard system is live and allows you to subscribe to our basic service. Sign up for a free account to explore the Zeitgeist around individual stocks and to set up basic alerts on up to 3 tickers.
For further insights, a system demonstration and access to our or upgrade to a paid subscription to set up a portfolio of tickers to follow and to enable additional alerts, or more advanced trading services there is more information here or contact us at firstname.lastname@example.org