We learned in a previous post that it is indeed possible to time the market using only volume with the TVO System. But what if you’re a more active swing trader looking for higher gains or just a little more excitement? Introducing the Heat Gauge System. So how does it work and how did it come about?
It started as something sitting right in front of my face and I still couldn’t see it. Or maybe I just didn’t want to. In technical trading, this happens quite often, which is part of the reason I steered away from chart patterns and conventional indicators. But the “safer ” world of backtesting and quantitative strategies is not without its pitfalls.
One of the reasons I developed TVO (Total Volume Oscillator), was so that I could in effect remove myself from my trading. Having a proven system at the helm steering you through the markets, allows you more time for other stuff, like writing the TVO Market Barometer blog for instance, which I started a little over a year ago.
Right off the bat after the first few posts, several day traders started giving me accolades for “calling the market,” which was a bit surprising considering most of my comments were not meant to be trading signals. I did post signals for the longer term TVO system (and still do), but others were apparently managing to profit in the short-term from my market volume analysis.
I was somewhat intrigued. Then it hit me that maybe there was something there that I could use as a basis for a system to capture short-term swing moves. And of course it would have to be low maintenance to suit my recovering trade-o-holic status.
After crunching some numbers I came up with an algorithm that measures volume like a thermometer or Heat Gauge (Hg stands for mercury after all), and the concept of trading it are simple… When it gets hot, get in the market and when it cools down, get out or go short. After several months and countless backtesting spreadsheets, the results came in and were astonishing to say the least. Here’s a chart of the initial equity curve.
The return on $10K in 15 years using SPY options (calls and puts) was just over 3 hundred thousand percent. Comparing that to the S&P may seem silly, but just in case you’re curious, SPY is that straight red line at bottom of the chart. Needless to say, I was extremely skeptical, so I continued to test the system. At the same time I started to trade it… Live, not paper trade (wouldn’t you?), posting signals in real-time.
And so the HG (Heat Gauge) was born, but the start was a bit rough. The first 3 trades were negative, and the big “W” correction in August and September resulted in substantial losses. A $50k investment would have been down -21% at one point. In October, however, there was speedy and complete recovery and as of 11/19/15 that same investment is now up 3% compared to SPY which was flat for the same time period. These are the actual trades.
If you’re up for the year and beating the S&P, it’s not normally the best time to start questioning your trading system. In fact, continuing to tweak the parameters can often be disastrous. I know this from experience. But how could the real returns be that far off from the backtests? What about my grateful group of day trading buddies online who made a killing?
It would have been easy to write it off as just a timing thing, but I felt like there was something else I was missing. After retesting all the data (always being careful not to “curve fit” the results), the obvious thing finally hit me… I was putting in long trades while TVO was in cash.
Oh, I was well aware that I was doing this, but I justified the trades thinking that the two were separate systems and should be treated as such. It can be common to have several independent strategies working at once. HG was born from the very same data as TVO, though, so I should have suspected that they might correlate.
And it turns out they do. TVO was in cash from 6/2 to 9/4. If the HG trades made during that time were omitted, the above return jumps to 13% and drawdown shrinks considerably to -14%. The answer was there all along and a momentary losing streak turned out to be the stroke of luck needed to help me see it.
By adding this parameter to the backtests, the amount of trades gets cut in half to around 150, which can be a good thing (remember less trading equals better quality of life). This does reduce our sample equity curve exponentially, but other important factors come to light and things shape up considerably.
In the new backtests, the success rate using SPY options is 65% (It goes up to 71% if trading the 2x leveraged ETF SSO), Profit Factor is 2.92, and account drawdown drops to -18%. That drawdown percentage may still seem like a lot for some folks, but our recovery factor is as high as 13 and the Ulcer Index, a formula used to measure how much you can stomach your drawdown, is a low 5.26%.
Since HG is a high probability strategy, what these other stats basically show is that it won’t take long to get your account back in the black from inevitable losses, and your chance of developing an ulcer worrying about it are greatly decreased.
And what about those other 150 trades that were lost after refining the system? I tested them separately and the success rate was below 60%, the typical low benchmark for sustained performance in this kind of strategy. Let’s take a look at the equity curve now.
The return here is just around 3700% and now you can actually see SPY (albeit still minuscule) in comparison. Once again the chart shows the growth of $10k over a 15 year period. For the sake of simplicity, adding to the position was omitted in the cases where a second signal was given. Also, no more than 10% of total capital was used for each trade. Average annual returns (CAGR) are about 26%. The following chart shows the return for each year vs. SPY.
HG never lost money and beat SPY with double-digit returns every year except ’01 (-1%) and ’12 (+3%). In both of those cases, the following year did more than make up the difference. And yes, the system made over 12% in ’08 while SPY lost over 37%.
Does it perform well in all types of markets? Are the returns consistently profitable year to year? These are things we look for in a robust trading system, and HG fits the bill… backward and forward.
Adding the market timing signals of TVO really helped to boost the performance of HG, but can you trade the two systems together? Yes, and the results get even more staggering… Using 70% of capital for TVO with SSO and 10% for HG with SPY options, the combined 15 year return was just over 16,000%, with average yearly returns at nearly 40%.
Why does it work so well? You may not want to believe it, but in today’s market, volume is really the only leading indicator. It’s the big thing under everyone’s nose that many folks overlook or refuse to see. Everything else, including price and thus every indicator related to it, is often manipulated by Market Makers and Wall Street professionals and therefore should be suspect. Follow volume and you’ll stay one step ahead of the herd.
If you’d like to see the complete backtesting reports, email me at firstname.lastname@example.org and I’ll send you a PDF.
Disclaimer: None of these signals or trades are to be considered investment advice. Always manage and limit risk especially when trading options.
Above post contains results from actual trades. Performance results on this website dated prior to September 2014 for TVO (prior to May 2015 for HG), including backtesting and trade history, are simulated. Please read our full disclaimer.