Bitcoin Investing: The Classic Managed Futures Approach
by Patrick Oberhaensli, on May 18, 2019 10:03:00 AM
Bitcoin made an accelerated upward-move in 2017 and reached its highest value near the 20’000 USD mark around mid-December of the same year. After getting to that absolute top, the bitcoin started to fall sharply until it entered the range within 3’000 to 4’000+ USD by the end of November 2018. It broke out of that range early April 2019 and reached 8'000 USD on May 14th. That’s not the first sharp up-and-down move we’ve experienced with Bitcoin and likely not the last one…
That’s why in this article we will have a first look at Bitcoin and the classic managed futures approach: these are strategies implemented with futures and using trend following systems to a large extent. The about 350 bn USD market of managed futures (it is likely larger than that) is mostly made of diversified trend-followers, but not only. When discussing trends, we need to be sure to mention that short-, medium-, and long-term- types of trends are the focus; it can also be a combination of these time horizons and moving averages and pattern recognitions are the main techniques used to identify trends.
Simple Moving Average
If we were to apply a 50 weeks simple moving average over the weekly values of Bitcoin in the last 2 years back from March 7th 2019, it would have led us to sell a Bitcoin position around 7’350 USD on May 20th 2018. Given that it was valued at 970 USD two years before, investing then meant a multiplication of the value by a factor of close to 7.5, and much more for those who went short instead of just selling that long position. If we were to use a range of 20 weeks instead of 50, we would get 4 buy and 4 sells signals with one buy and sell leading to a smaller loss: the first sell would have come at about 9’250 USD instead of the 7’350 USD.
We see here that identifying a trend – the time needed associated to that recognition - will certainly lead to entering the trend late... and leaving it too late as well, that is when the trend doesn’t exist anymore or it has already reversed. This is the common drawback of trend following strategies. Even though in that case the faster moving average (the one with fewer weeks look back) has led to higher gains, the first sells signals came really too late. It should also be noted that a trend following strategy will lose not-too-much money when the identified trend was actually false (in other words, it could simply be a trading range).
Bitcoin Behavioural Trends
The key question: Does Bitcoin move in trends? The answer is apparently yes. However, the problem is that one could get trapped by the analysis of a specific past period which won't necessarily be representative of future behavior. The last few years have been quite particular with an “exponential” move upward in the five dollar digits followed by a move downward that was very strong and much less “smooth” (not a smooth curve which would be less beneficial for trend followers). Nevertheless, the results are (much) better than with a buy-and-hold approach. The long-only approach is not adapted for such an instrument given the numerous sharp moves in both directions (that's volatility). This leads to the likely conclusion that the bitcoin market isn’t even slightly efficient or "weakly efficient" when referring to the Efficient Market Hypothesis. Would it be weakly efficient then price and volume information from the past would not allow for long(er) term abnormally high risk-adjusted returns.
Bitcoin Pricing Trends
The latest price range observed for Bitcoin was graphically easy to identify and suggested the use of non-trending techniques to generate a profit. One should be cautious, classic approaches in that context could potentially (if not likely) have signalled wrong overbought and oversold situations. The break-out early April was vigorous and is consistent with the fact that the trading range had lasted for several months but still of astonishingly strong amplitude. Trend-followers would see here a trend-up – keeping in mind that it still could have been a false signal also known as a “whipsaw” signal at the beginning.
Finally, the discussed rules to manage Bitcoin-investments can be modelled and programmed as Quant rules. The models used in this case are quantitative models, also called quantitative technical analysis (as opposed to qualitative technical analysis).
We have shown here that a systematic approach in trend-following, if modelled properly, should sharply improve a purely passive approach. In our opinion, there are even “smarter” ways to apprehend the movements in Bitcoin but one should never forget that it remains a non-liquid instrument despite its (still) dominant market share among the cryptocurrencies: the whole market is small. Moreover, the futures market in Bitcoin is still very young and dealing with large(r) sizes is unsurprisingly an issue. These questions have led to us to examine a quite different approach of investing in Bitcoin.
EVOLIDS FINANCE LLC, Disclaimer:
- This content is not intended to be a solicitation nor an offer
- The preparation of the information provided herein is done with a high level of care. Nevertheless, errors are possible
Are you interested in a serious Bitcoin investment strategy?