Unravelling the Financial Markets with data analytics

Written by admin

July 14, 2022

 

“I love quotes… but in the end, knowledge has to be converted to action or it’s worthless.” – Tony Robbins 

So, whilst we have been very quiet on the online front it’s been a hectic year for the team at Trusted Data, focused on our new pet analytics project….

In Jan 2021, in the midst of a new Covid lock-down, with client face-to-face interaction restricted, our team brainstormed areas where our analytics expertise could deliver actionable intelligence.

One of the areas, long discussed in industry, has been how to unravel the mysteries of the stock market. Many we spoke with said it could not be done, even with data analytics and that constantly shifting market dynamics made it impossible to predict stock market movements, especially intra-day/week moves.

Not ones to say no, we set about testing various theories and methodologies, targeting the formulation of at least 1 coherent trading strategy, which we could then deploy in a live environment.

We thought that we struck gold on several occasions in the first 2-3 months and whilst the results were decent, we were not satisfied that the trading strategies created were robust enough to be successfully applied on a long term basis.

Being an organisation that does not like to speculate on what if’s / but’s and maybe’s, after 1 month of extensive back-testing, we deployed funds to the market to test our mettle. The story so far is summarised below and may help any interested and aspiring investors and day-traders.

 

Considerations

“Based on my own personal experience – both as an investor in recent years and an expert witness in years past – rarely do more than three or four variables really count. Everything else is noise” – Martin Whitman

 

  • We avoided going down the sentiment analysis / media route as we were not confident that it could be modelled adequately and provide the explanatory power we sought to predict market movements.
  • Whilst initially focusing on individual company stocks, we found more stability when testing leading indices.
  • Investing was of limited interest to us to begin with, during back-testing, as we wanted to test multiple short term trades to generate a decent sample size in a confined time period
  • We saw crypto as a bubble quite early on and decided to stay away from it, though the temptation to short BTC on 9/10 November was huge, when we picked up the reversal!
  • Our focus thus shifted to back-testing trades on leading indices on a 1 – 5 day time horizon, via spread-betting
  • We looked at multiple market indicators and quickly confused ourselves before taking a step back and focusing on seeking out a simpler approach to find success

 

Approach and initial insights

 “No wise pilot, no matter how great his talent and experience, fails to use his checklist” – Charlie Munger

 

  • Indices had unique attributes in terms of rhythm and daily trading ranges (hi-lo) and we developed a robust understanding on this very early on to project intra-day index movements
  • A number of indicators looked very potent when trying to repaint potential trades against them but this was misleading and we quickly discarded any obsession on repainting
  • Indicators cannot always be universally applied across equities. After running an optimisation test on Parabolic SAR for example, we found that the SAR settings needed to be customised against each index we were trading
  • Pivot points provided very useful reference points for targeting where to open potential long / short positions and if untested on any given day, could formulate potential price targets for the coming 1-3 weeks
  • Pre-defined support / resistance levels were interesting referral points but past behaviours would not necessarily inform future support / resistance levels, especially if you are trading potent market breakouts
  • One of our first trading strategies uncovered strong relationships when an index was in an intra-day trend and our strategy was to trade part of this trend. Whilst the strategy had a +96% average success rate, it required regular updating (due to fluctuations in index values and their trading ranges) and would also cap returns to 80-100% p.a. We therefore discarded its’ use after approx 4–5 months.
  • Our preferred trading intervals to enter and close trades became 15 mins, 1 hour, 4 hours and Daily, though we utilised weekly trends as a backdrop to assess potential thresholds (hi-lo) 
  • We then assessed and modelled other indicators to uncover potential breakouts in the market and spent countless hours also observing minute-by-minute candles to learn the potential components and behaviours of daily market reversals

 

Mistakes and Learning outcomes 

“The intelligent investor is a realist who sells to optimists and buys from pessimists” – Benjamin Graham 

  • Leveraged trading, especially spread-betting is not for the faint-hearted and should be avoided unless you have solid confidence in the market and your trading methodology
  • Do not trade on sentiment and when emotional, please step away
  • If your trading strategy works, focus on that as opposed to assumptions on when you get a ‘good’ price
  • Do not trade because you feel you have to on any given day, the best trading outcomes can come from constraining trading to even just 1-2 days per month – the market will always give you big opportunities each month -we learnt this the hard way in month 1!
  • Avoid setting yourself revenue / return targets as they are likely to force you into unnecessary trades
  • Averaging down is a useful approach for risk management but only when carefully managed and when you have confidence on where price bottoms will be. Do not chase the market up or down based simply on trying to find a top / bottom.
  • Try not to let media articles influence your buy / sell decisions 
  • Just because you have available reserve equity, doesn’t mean you deploy it heavily to active trades (our preference has been to utilise between 15-25% of total equity on live deals + market orders at any given time)
  • It can be good to become somewhat of an expert on just a handful of equities but at the same time, we found our trading strategies could be deployed across multiple equity classes and our portfolio of trades diversified in early 2022, first focusing on individual stocks and then moving to commodities, though the margin requirement can to open trades on individual stocks or some commodities is often proportionately far greater than index trading.

  

Results 

“You make most of your money in a bear market, you just don’t realize it at the time.” – Shelby Cullom Davis 

As our understanding of the markets developed, coupled with the specific equities we traded and refinement to our trading strategies (resulting in the development of our own trading platform) our confidence and results grew significantly….  One bugbear of ours has been the online brokerage platform we have been using to date, and their fee structure, which we feel is sub-optimal. This has resulted in a +13% shrinkage from gross-to-net results in 2021 and 18% gross-to-net shrinkage 2022 YTD. 

Net Results adjusted to annualised % format (quarterly net profit * 4 divided by the total available equity at the start of that quarter) 

  • Q1 2021 – 9.9%           The poorly executed short on Dow Jones really hurt in month 1! Greatest learning lesson off the bat for us.
  • Q2 2021 – 32.9%         Making strides … long positions in FTSE in the 6500 range played a pivotal role in the move upwards
  • Q3 2021 – 104.5%       More like it – summer volatility opened up big trading channels, wrapped up by excellent Nikkei shorts in Sept
  • Q4 2021 – 205.7%       November projection of 2022 market crash…Short in early December, long on Santa rally and Short in wk 4 Dec
  • Q1 2022 – 110.1%       Poor entry points on Nasdaq and HSI in the midst of a geopolitical crisis hurt…Our second big learning lesson
  • Q2 2022 – 252%           Shorts on FTSE 100, Energy, Commodities and long positions in Nasdaq 100 in June to wrap off a strong quarter

 

What next? 

Trusted Data started as a data analytics company solving unique industry problems, by building one-off analytical / Ai solutions for clients to address a specific need / problem statement.

By utilising latent time in the midst of Covid, we pivoted and focused intensely on a new problem statement in the form of the financial markets, tackling it head on, not in a theoretical context but in a live environment with real funds and real emotion.

Whilst we retain an acute interest in delivering client engagements, particularly in the domain of predictive and optimisation analytics, the move towards the financial markets has sparked phenomenal interest for our Company and potentially offers the next evolution in our business, not only focused on developing solutions for external clients but also for our internal team to deploy to the markets across commodities, FX, indexes and stocks.

If you would like to find out more about our Company and / or how we consider, design and deploy analytical solutions to solve unique problems, please feel free to get in touch (info@trytrusted.com).

You May Also Like..

What you can or cannot forecast

What you can or cannot forecast

What you can or cannot forecast   Forecasting is required in many situations: deciding whether to build another...

Commodities are a Bubble ???

Commodities are a Bubble ???

Commodities: Risk or Opportunity?   The Trusted Data team have been monitoring the commodity market quite a lot...

0 Comments

Submit a Comment

Your email address will not be published.