As regular readers may or may not recall, forward testing of the Happy Hunter Price Action Trading System started way back on February 2018. And originally, the plan was to run the forward test for a year.
The forward testing phase therefore still actually has a month to go before it technically ends.
However, I think it’s already quite apparent that the system is very robust, even though GBP/NZD’s volatility levels and price action during the forward testing period weren’t exactly within historical norms, especially during the first half of the testing period.
And remember, I crafted the system based on the assumption that GBP/NZD’s price action would continue to respect historical norms.
Also, do keep in mind that ALL trades assume a fixed 12-pip spread, which is a rather substantial hurdle to overcome, especially for a day-trading system.
Anyhow, I’m already veering off course…
Getting back on topic, the purpose of today’s write-up is to present the final and total results of the system, as well as give a heads up on my plans for the system.
I’ll also show you how the system would have fared if we lived in a fantasy, zero-spread environment. Don’t be too surprised with the yuuuuge difference in performance, though.
Of course, I’d also like to take this opportunity to reminisce a bit on the Happy Hunter’s evolution and my personal journey with it.
Anyhow, here are today’s topics:
- The Saga Of The Happy Hunter
- Final Test Results (Fixed 12-Pip Spread)
- Final Test Results (Zero Spread Fantasy Universe)
- Going Forward
The Saga Of The Happy Hunter
For those who don’t know, the catalyst that sparked the development of the Happy Hunter Price Action Trading System was the realization that I’ve never really gone out of my comfort zone as a trader.
I therefore resolved to do just that in My 2018 Trading Resolution by devoting myself to creating a mechanical trading system from scratch.
And to push the boundaries of my comfort zone even further, I challenged myself to create a trading system that must be the complete opposite of my current trading framework.
The trading system must therefore be:
- Purely mechanical
- Not dependent on fundies at all
- Reactive not predictive
- Designed specifically for day-trading
I also defined my trading system parameters, which can be summarized as follows:
- I will only risk 0.5% of my capital per trade
- I will only use the 1-hour timeframe
- I will only trade GBP/NZD
- I will only use stop and market orders for my entries; limit orders are forbidden
Of course, I also naturally defined my goals for the system, which are as follows:
- My peak-to-trough drawdown must not exceed 20%
- My reward-to-risk ratio must be at 1:1 or better
- I must open at least one new trade per day
- No losing year (Easy Mode)
- No losing month (Normal Mode)
I also defined my bonus goals (and inserted some joke goals):
- A win rate of 50% or better
- No losing week (Hard Mode)
- No losing day (Harder Than Hard Mode )
- No losing hour (This Is A Joke Mode)
- No losing minute (Holy Grail Mode)
- No losing second (Delusional Fantasy Mode)
And to narrow my focus and have a starting point, I decided to use these two trading principles/axioms as the core of my trading system:
- A period of consolidation is usually followed by a period of expansion
- All directional movements have an exhaustion point (i.e. all trends end)
Given the above trading principles/axioms, and taking my objective to create a purely mechanical trading system into account, I decided to focus on these three candlestick patterns since they can be mathematically defined:
- Higher Highs, High Lows (HHHL Patterns)
- Lower Highs, Lower Lows (LHLL Patterns)
- Inside Bars (IB Patterns)
And with that, the founding principles and objectives of the Happy Hunter Trading System were set.
About a week after I wrote my 2018 trading resolution (and thanks to some hints from an experienced trading system developer), I finally finished the first prototype of the Happy Hunter.
And while Version 1.0 looked good on paper, that’s only because spread was not incorporated into the backtest results.
Once a fixed 12-pip spread was taken into account, the system quickly became a loser.
However, the one great lesson from backtesting Version 1.0 that we learned is that using HHHL and LHLL patterns as entry triggers gives us a statistical edge over the market since the hit rate is above 50%.
The goal then was to find a way to increase profits in order beat the spread problem. And the two most direct solutions are to either:
- Find a way to let our profits run
- Find the statistically optimum fixed profit target
And so the beginnings of the Fixed TP Variant and Trailing Variant began to form.
The first iteration of the Trailing Variant was finished first and was presented as Version 1.1. And shortly after, the first attempt of what would become the Fixed TP Variant came in the form of Version 2.0.
Version 2.0 was also the first version that used what I call the “multiple entry type” approach.
And initially, the “multiple entry type” approach was used to differentiate how the HHHL and LHLL patterns were used since HHHL and LHLL patterns under Version 1.0 and Version 1.1 were used to define potential exhaustion points and/or reversals, and Version 2.0 began using HHHL and LHLL patterns as breakout patterns essentially.
Anyhow, building on the lessons learned from the earlier versions, Version 2.1, 2.2, and 2.3 were later simultaneously rolled out.
Version 2.1 was essentially the second step in the evolution of what would be the Fixed TP Variant, while Version 2.2 was the evolved form of what would later become the Trailing Variant.
As for Version 2.3, it was a hybrid of Versions 2.1 and 2.2.
I ultimately gave up on the hybrid version and focused development of what are now known as the Fixed TP and Trailing Variants.
However, the major milestone that was reached when developing Version 2.1, 2.2, and 2.3 was the implementation of rules to move the stop loss (SL) to a reduced stop loss (RSL) in accordance with the objective to preserve capital by limiting risk exposure when the trade is going our way.
It’s worth pointing out, though, that rules for using the IB pattern were not yet laid down.
That would come later when Version 3.0 was rolled out. And with the IB entry rules in place, the Happy Hunter Price Action Trading System was finally complete. Well, Version 3.0 was the first completed version anyway.
Incidentally, the forward testing phase began after Version 3.0 was developed.
Unfortunately, February 2018 was a difficult month for the trading system since trading conditions weren’t particularly favorable.
But on a happier note, the Fixed TP Variant showed how well designed it was since it closed out the month with gains despite the poor trading conditions. Not really much of surprise, though, since the Fixed TP Variant’s underlying philosophy is to take profits at a statistically probable price levels.
And on a painful but constructive note, the unfavorable trading conditions in February exposed the Trailing Variant’s major design flaw, namely its tendency to give away too much of its gains during prolonged periods of choppy, sideways-moving price action.
And so work on a solution to address the Trailing Variant’s flaw began and Version 3.1 was later rolled out.
However, Version 3.1 also had its flaw, namely that too many profitable trades were missed. And so the entry rules were overhauled to be less restrictive and Version 4.0 was later rolled out.
And so far, Version 4.0 has proved itself to be more than adequate for the job of capturing gains from the market, despite the not-too-favorable trading conditions this year.
And that gives me the perfect opportunity to segway to…
Final Test Results (Fixed 12-Pip Spread)
Both Variants are printing very impressive numbers after 11 months of forward testing, with the Fixed TP Variant up by an amazing 27.39% and the Trailing Variant up by an astounding 54.87%.
And again, do keep in mind that each and every trade assumes a fixed 12-pip spread.
Anyhow, as you can see above, the Fixed TP Variant’s equity curve has always had a slight upward tilt, while the Trailing Variant’s equity curve only achieved liftoff during the later half of the year.
And that’s because trading conditions were rather poor (and even downright terrible) from February until July.
July was particularly soul-crushing. Just look at the equity curves of the two Variants back then. That’s a nine-day, non-stop, losing streak that wiped out 13 days worth of gains, yo!
As to why the Fixed TP Variant had a tough time while the Trailing Variant had a terrible time during the first half of the forward testing period, that’s because GBP/NZD was consolidating tightly back then.
You can even see this on the daily chart.
And to give y’all some context, compare GBP/NZD’s price action during the first half of the forward testing period with how GBP/NZD’s price action historically played out.
Really puts things into perspective, huh?
Anyhow, trading conditions finally began to improve when GBP/NZD began to move up back in August in order to try and stage a topside breakout from that ascending triangle.
And when the pair finally successfully staged a topside breakout in October, Christmas came early for the Trailing Variant since it was able to open several trades in the direction of the uptrend and those trades were able to ride large chunks of the breakout move, which is why the Trailing Variant was able to harvest a very bountiful 25.02% in gains back in October.
The topside breakout ultimately failed, and the pair went back inside the triangle, only to break to the downside, which meant more volatility and directional movement, so both Variants continued to score wins in November and December.
But if you’re wondering how the system would have fared if we lived in a fantasy, zero-spread world, then lucky you since that’s what the next part is all about.
Final Test Results (Zero Spread Fantasy Universe)
Yes, those numbers are correct. Those numbers were double checked and triple checked by yours truly.
So yes, if spread didn’t exist, then the Fixed TP Variant would be up by a mind-numbing 186.00% after 11 months, while the Trailing Variant will be up by an astronomical 204.80%.
And no, there’s no compounding effect, and yes, the capital base is the $10,000 principal.
But before you curse and shake your fist at your broker, I should stress that the difference in performance is partly due to my strong emphasis on percentages and very strict cost control.
As you may or may not have noticed, I don’t really gauge my system’s performance based on the amount of pips gathered – my system’s returns are always on a percentage basis, which is only logical since the lot size used varies, given that the stop loss also varies depending on the value of ATRH.
If ATRH is bigger, then the stop loss in pips is bigger, but the full loss from a single trade will still amount to 0.50%.
If ATRH is small, then the stop loss in pips will be smaller, but again, the full loss will still amount to only 0.50% of my capital base.
The variable stop loss was meant to make the system self-adapt (so to speak) to changing market conditions.
Basically, if trading conditions tighten, then ATRH shrinks and the system can afford to open positions with larger lot sizes, enabling the system to exploit the opportunity when volatility picks up again.
Conversely, if volatility picks up, then ATRH increases and the system theoretically becomes less likely to get stopped out by market noise.
However, that’s where spread becomes a major obstacle.
You see, the system’s stop loss in pips is variable, but the fixed 12-pip spread is, well, fixed. And if you want to understand more, then you can read my write-up on The Problem With Spreads.
Anyhow, the bottom line here is that if we practice strict cost control and take spread properly into account, then spread lowers our reward to risk ratio by inflating our risk while simultaneously dampening our profit potential.
However, that’s not the end of it. Since our risk is inflated by the spread amount, we have to compensate by lowering our lot size. Otherwise, we end up risking more than we want to.
And reducing the lot size also hurts our profit potential again, so it’s a double whammy.
And as a concrete example, the Fixed TP Variant’s average gain from a full win under a 12-pip spread environment is around 0.43% of the capital base or $43. However, a full loss will always be around 0.50% or $50. Do note that you stand to gain less than you lose.
But in a zero-spread environment, the Fixed TP Variant’s average gain from a full win would be 0.70% or $70, which is 0.27% or $27 more when compared to the average gain in a 12-pip spread environment. But at the same time, a full loss will still be around 0.50% or $50, so in this case, you stand to gain more than you lose.
Now also take into consideration that there were 582 full wins for the Fixed TP Variant. It quickly adds up, huh? By the way, that’s where the 157.14% or $15,714 in profits went.
The forward testing phase is finally over. So what now? Well, I will no longer be providing daily updates, for one. The rules of Version 4.0 of the system will always be publicly available for all to study (and copy), though.
But with regard to the trading system itself, well, the system is still actually inherently flawed, given the lack of filters, which translates to painful losing streaks during prolonged periods of choppy, sideways price action, coupled with relatively tight volatility.
Another flaw that’s specific to the Fixed TP Variant is the Variant’s poor reward to risk ratio due to the spread.
There’s therefore still room for improvement.
And as regular readers may know, Version 5.0 and Version 6.0 were in the works in order to rectify those flaws.
Both Versions 5.0 and 6.0 are already partly complete, but as I stated back in August, the way things are going with my tasks, trading, and life in general, I probably won’t have time to do so in the foreseeable future, so I’m scrapping any future developments and Version 4.0 will very likely be the last version of the Happy Hunter Trading System.
I may eventually complete development of Versions 5.0 and 6.0, but I don’t know if I will be sharing them publicly the way I did with Version 4.0 and its predecessors.
Anyhow, hopefully y’all were also able to profit from my journey, not just in term of money (assuming you studied and copied my system), but also with regard to your own growth as traders.
Maybe some of you were even inspired to create your own Variants of the Happy Hunter. It’s even possible that one (or more) of you were inspired to create a totally different trading system of your own design that’s more profitable than the Happy Hunter.
In any case, we had a pretty good run, didn’t we?