We are releasing a new EA, "Wave Rider Kinjiro"! Can it cover the slump of the previous work, SuperRecovery?!

I was endlessly torn whether to release a new work while Super Recovery was having issues, but since there is still a possibility that the new work could cover Super Recovery, I will bravely publish the new work.

The new work is “Friend of the Waves: Kinjiro” created to compensate for Super Recovery losses after the launch of Super Recovery.

About three months have passed since development, and since there is a prospect for release, I decided to make it public.

Will the new work cover the prolonged slump in Super Recovery where profits were not being made? I am full of both expectations and anxiety about this.

Super Recovery is still not performing well, and when considering the causes, it relates to the rapid increase in capital flowing into the USD/JPY market from 2022 onward—the “big bull market” where the USD/JPY surged from around 110 to over 150 due to expanded volume.

Logics and parameters optimized for a big bull market cannot respond quickly to long periods of ranging markets with trend-following MA logic, as seen last year.

The biggest feature of Super Recovery for bull markets is that by employing a pyramiding method during such sharp, extreme volatility, the backtest for 2022 was outstanding.

The figure below shows Super Recovery’s test results for the full year 2022, where 100,000 yen grew to 50 million yen.

However, in 2023, the figure shows that 100,000 yen only increased to 600,000 yen.

We believe these extreme differences in total profit and loss are largely due to the magnitude and duration of market trends.

GuppyEA, the creator of Super Recovery, uses the same 4-hour logic, yet its profit from 2022 to 2023 grew to around 100 million yen from 100,000 yen, with win rates of 100% for shorts and 96% for longs, showing strong performance last year as well.

Trading frequency over two years has been relatively low at 60 instances, but it has produced steady profits in 2023 as well.

The idea behind developing Super Recovery—the method that anticipated that the 2022 bull market would continue into 2023—also became its biggest weakness, causing the range-bound market in 2023 to trigger frequent stop-outs when trends were small.

What will happen this year is anyone’s guess...

Especially around the end of last year, government interventions and US rate hikes caused the yen to move unexpectedly, and fears of intervention led to rapid declines, revealing that the MA’s strength in steady responses did not apply to Super Recovery’s weakness in initial reaction.

As of today (February 1, 2024), the USD/JPY is oscillating between expectations of a Federal Reserve rate cut and a US recession concern and Japan’s rate hikes, creating wild swings.

This is precisely where a long-range MA-based pyramiding method struggles in such wide-ranging markets.

Currently, Super Recovery’s logic centers on the most recent high and low of the 4-hour timeframe, but in recent markets, after the 4-hour high is broken, prices push down hard and then rise sharply again, making it difficult for MA to keep up—a major reason for the losses.

From the start, I anticipated some range-bound weakness, but when the range-bound market lasts this long, a trend-following MA-based pyramiding approach does not perform well.

In short, Super Recovery is now in a trend-waiting state, and the question is how long it can endure losses.

Specifically, if prices rise past 152 yen and push toward 160–170, Super Recovery could yield large profits similar to 2022.

Or, if prices break below 145 yen and head toward 130–120, I believe the EA could also deliver substantial profits.

Until markets become such extreme and clear, profit expansion may be difficult.

For high-leverage or high-risk traders, this is a painful caution for Super Recovery at present.

Another characteristic of Super Recovery lies in its risk-reward approach.

It aims for small risks and large rewards. In the current version, the stop loss is 204 pips, while the take profit is 312 pips, which is relatively large.

In reality, many cases involve trailing or MA settlements to secure smaller profits, but I believe a more granular profit-taking logic is necessary.

In the new work, “Friend of the Waves: Kinjiro,” I have increased the number of trades to about 7,000 over two years by reducing maximum profit per trade and by trimming take-profit and trailing features.

In other words, the new EA is a move from swing-type Super Recovery to a scalping-type Kinjiro.

Of course, depending on timing, Kinjiro may carry positions to the next day or week.

Super Recovery conducts around 700 trades over two years, while the new Kinjiro does about ten times that, around 7,000 trades.

Moreover, Kinjiro trades only on Wednesday, Thursday, and Friday within a single week, reducing activity to three days out of five, yet the total number of trades is ten times higher, indicating a highly active trading frequency.

In short, the idea for Kinjiro arose from lessons learned from Super Recovery’s underperformance.

Here is the MT5 real-time backtest result for Kinjiro over the most recent two years.

With a fixed lot of 0.46, 100,000 yen grew to over 16,000,000 yen in two years.

In a domestic brokerage account, that would be roughly 1,600,000,000 yen.

I will explain in order, but before that, from the failures of Super Recovery, one major lesson is that one of the primary reasons for large losses in range-bound markets is the auto-lot feature.

Therefore Kinjiro uses a fixed lot size this time.

1. Test period: January 1, 2022 to January 24, 2024 (2 years)
2. Longs and shorts combined up to 21 positions as the upper limit
3. Stop loss
4. Fixed lot size
5. Trailing start
6. Trading starts on Wednesdays
7. Thursdays
8. Only three days per week on Fridays
9. Tick-based historical data
10. PF 1.26 RF 4.88
11. 7,085 entries over two years
12. Drawdown 22% is acceptable
13. The win rate is very low because when a position closes, it always starts from hedged positions, so several stop-outs occur at each closure

The logic that adds positions only in the direction with unrealized gains creates this relatively low win rate.

And my interest is not win rate but how much total profit can be generated in a short period, so I tend to favor high-profit, low-win EA.

Also, while Super Recovery uses trend-based MA on a 4-hour basis, Kinjiro builds its logic purely on price action without using indicators, oscillators, or candlesticks, which is a major difference from the new work.

Backtests show a maximum of 0.46 lots.

With high leverage and 100,000 yen initial capital, 0.46 lots × 21 positions equals 9.66 lots as the maximum holding lot size.

Kinjiro’s main feature is that there are often unrealized gains, allowing many positions to be held even with limited capital.

Of course, this is with high leverage.

In domestic accounts, it would be prudent to keep the lot size around one-tenth, or increase initial investment to around 1,000,000 yen.

Super Recovery’s main entry logic uses the Recovery indicator and only entries in the same direction on the 6-minute timeframe after a confirming 4-hour MA trend signal, whereas Kinjiro centers its logic on price action, making timeframes irrelevant and simply using price in pips to implement a grid method.

What does this mean? The grid is the default grid displayed when MT5 is installed.

It refers to a method of repeating entries at equal intervals.

The grid is composed of vertical and horizontal dashed lines in a grid pattern.

As shown in the figure, by default the grid is displayed in 22 pips increments, depending on the timeframe.

The above shows Kinjiro’s trading history for this week.

The grid is set at 9 pips.

In other words, it is a method of continuously adding positions in the direction with unrealized gains by 9 pips at a time, pyramiding.

The trading history during large downward moves on the chart illustrates Kinjiro’s characteristics, and this capacity to quickly enter positions at times of important statements or data releases is because the logic focuses solely on price action.

Using the grid to add positions at 9 pips increments, pyramiding only in the direction with unrealized gains.

In short, a scalping-based pyramiding method.

Regardless of the exact logic, my basic principle is trend-following and adding in the direction of the trend, so I again chose a pyramiding approach for Kinjiro as with Super Recovery.

That is the only point where Kinjiro and Super Recovery share a commonality.

Why insist on pyramiding? Because of painful experiences with mean-reversion averaging-down (martingale) strategies.

Mean-reversion averaging-down strategy adds positions in the direction of losses, which is the essence of averaging-down and martingale.

When adding positions in the direction of unrealized gains, it is called pyramiding.

Kinjiro starts from hedged positions immediately after the EA is installed.

Trading begins with a 1:1 ratio of long and short positions.

When the EA starts, both long and short entries occur at the same time, fixing both potential profits and losses initially.

If the price rises by 9 pips, one long position is added.

Each time the price rises by 9 pips thereafter, one more long position is pyramided.

If the market reverses and falls 9 pips below the initial short position and the short positions become unrealized gains, one short position is added.

At that moment, all the initial long positions become unrealized losses, and the previously unrealized short positions turn into unrealized gains.

Thereafter, entries continue only in the direction with unrealized gains, up to a maximum of 21 positions.

If, for example, at the time of reaching the maximum 21 positions there are 3 shorts and 18 longs and the price continues to rise, large profits occur while the 3 short positions become unrealized losses.

If the market moves opposite to many positions, it results in a stop-out.

The imbalance between unrealized losses and gains, and the trend continuing in the larger-position direction, leads to substantial profits and an expanding total profit and loss.

There was a drop of about 200 pips in the shadowed area in the image above.

Let's zoom in on the same scene below.

The magnified hourly USD/JPY movement for February 1–2 of this year shows entries repeated at 9-pip intervals in the direction with unrealized gains.

As shown above, Kinjiro pyramids positions at 9 pips increments, but large price movements cause slippage and entries are not made at perfectly equal intervals.

As a rough guide, the price-action-based position acquisition method uses 9-pip grid only.

On the left side of the figure, longs are stopped out; in the center to the right, large profits are produced from shorts.

As a result, this day ends with a substantial net gain because gains exceed stop-outs.

February 2’s trading history is shown here.

With an initial margin of 1,000,000 yen, daily profit was 114%, making total profit and loss 114,862 yen.

Thus, regardless of which time frame the EA is used on, Kinjiro’s unique logic remains unaffected.

Whether installed on a 1-minute or a daily chart, entries occur at 9-pip increments.

I will write the continuation tomorrow.

Today is February 3, calendar date changed.

I have just eaten an ehomaki.

According to Google AI...

The lucky direction changes every year and is determined by the year's stem-branch (jūnkan). The year 2024 corresponds to the stem-branch “Wood” (Kou), and is determined accordingly. The origin of ehomaki traces back to the Edo period, though there are multiple theories about its origin.

...that’s what they say.

By the way, when I woke up this morning I was startled by the dollar-yen moves.

I expected declines, but it turned into a rise instead!

The USD/JPY soared.

Shorts were caught out, weren’t they!

When shorts get caught, it means a large number of short positions were accumulated as people anticipated a drop, and then some factor caused a sharp rebound that compels short sellers to cover, driving prices higher than expected.

By the way, I won’t make tomorrow’s forecast, so I was flat today.

Shorts probably had a lot of stop-loss hits, leading to a forced buyback and a significant rise.

So, how did Kinjiro fare in this sudden move, as I wrote in yesterday’s article?

In this post-news surge, Kinjiro quickly added Long positions and took profits aggressively.

If you zoom in, it looks like this.

From February 2 alone, 1,000,000 yen of initial capital grew to 1,310,000 yen, giving a total profit of 310,000 yen.

That is a daily return of 131%.

After the positions were closed, Long positions were reacquired, so there are some positions carried over at the end of the test, but since Longs and Shorts are in a hedge, there’s no problem at the start of next week.

The final trades in the chart are buy sell buy sell.

On the far right, it says end of test, indicating the positions currently held.

In any case, assets increased by 131% in one night.

In reaction to the jobs data, such a sharp price move occurred, but the grid-type scalping EA “Kinjiro” excels in this kind of market.

For such drastic price fluctuations, trend tools like MA cannot react quickly enough.

Their reaction is too slow.

The fastest reaction is a tick, so nothing that is faster than price action exists in the market.

Kinjiro’s logic is built on price action alone, immediately incorporating the typical principle that news releases and key speakers move the price.

Kinjiro repeats entries purely based on price action, so the strength of the recent jobs data significantly amplified its advantage.

By leveraging these features, I considered making an EA that only operates around big indicators, but the trading frequency would be too low to be satisfying. What do you think?

It also depends on how buyers use it, right?

Now, I will explain Kinjiro’s drawbacks.

Mean-reversion averaging-down methods, which are common in the market, are strong in range markets and accumulate small profits, but Kinjiro’s scalping and pyramiding approach makes it easier to incur losses during tight ranges. (Referred to as Scalpyla EA)

Using only a 9-pip grid to add positions in the direction of unrealized gains can cause losses in narrow ranging markets because trends tend to reverse.

Because the maximum number of positions is limited to 21, if the losses become large, it hits the stop.

Let’s look at the most stagnant two years.

Total profit and loss surpasses 1,670,000 yen from an initial 100,000 yen over two years, equivalent to about 16,000%—quite good for a simple compounding approach.

The two-year period includes a maximum slump of three months.

Both Super Recovery and Kinjiro experience slumps for months, though Kinjiro shows less disparity between 2022 and 2023, and pyramiding means big profits in major markets remain possible.

As shown above, 2023 also included a three-month stagnation period, but in the fourth month it recovered more than half of its losses.

During the three stagnant months, profits gradually declined due to stop-outs and take-profits, but there has been no major stagnation since then, though minor fluctuations persist, with the biggest drawback being frequent stop-outs.

Nevertheless, assets have continued to increase gradually, so there is reason for future optimism.

Now, finally, I will explain the external panel.

If you enlarge it, it looks like this.

The top Buy Sell shows the current number of positions.

Below that is the total Pips for longs and shorts combined.

Below is the amount display.

Below are buttons to stop and start the EA. Pressing stop will halt new entries, but trailing stops, stops, and take profits on currently held positions will still function.

“EA Stop” means do not take new positions.

At the bottom is the “Close All” button, which closes all long and short positions.

Therefore, if you press EA Stop before this button, you will be net flat and will not begin new hedged positions.

Conversely, you can keep the EA Stop pressed most of the time and only Start EA when major indicators are released or a trend is expected, etc.—that’s up to how the user uses it.

How you use it depends on the user.

With that, I’ll finish explaining the new Kinjiro EA.

Actually, there is another completely different EA completed, and I will introduce it in detail in the next article.

Today, I’ll briefly introduce the second new work.

The next new work, under a simple one-year compounding operation with risk-reward of 1:1 and a stop loss of 40 pips and take profit of 40 pips, shows an increase from 1,000,000 yen to 10,000,000 yen in one year.

With compounding, 100,000 yen grows to 27,000,000 yen in one year.

This is not an improvement of Super Recovery, but a completely different logic.

Entry for the Low-risk version uses the 1-hour MA.

The High-risk version uses the 10-minute MA.

Kinjiro is entirely different in logic as well.

The EA name is “Super Winners.”

Two sets are provided: a Low-risk set file and a High-risk set file.

One EA, two set files installed together.

Backtests with 100% history quality show that the High Risk version grew 100,000 yen to over 200 million yen in two years.

Of course, such test results are based on idealized EA limits, so in reality risk settings should be far lower, but these are illustrative possibilities.

The following two-year backtests are for the second new work “Super Winners.”

In the next article, I will reveal details about the second new work.

Also in the next article, I will disclose a special price for Super Recovery purchasers and the regular price.

Stay tuned!