The Finance Book Club

They are probably hanging out with the guy who bought Winzip.

Ok, this just keeps cluttering up my to-do list so let's get it out of the way. It's only been 7 months.

I said I would do a video. I don't think I'm a video kind of guy. I'm more of a type-it-out kind of guy. Maybe some day I'll 'upgrade' to videos. Not today.

The issue is the LAST KISS method described in 'Naked Forex' by Nekritin and Walters.

Here are the rules again just in case you don't have a seven month photographic memory.


As I mentioned before, there are no actual live trades to consult on the book website. We are on our own here, as always.

The main problem seems to be that the 'zones' described by a ranging price 'box' are poorly defined by the authors. It's really, really important to all their trading methods too. All they say about the 'zone' is that price needs to make two contacts with it, but they don't say how thick it is, or how long we stick with it, how it gets redrawn, how much tolerance it should have when it's violated. We can sort of extrapolate based on the diagrams in the book, but this, for me, with my level of understanding of 'the game' is nearly a deal breaker. If you can't specify with measurements, in pips or points, using numbers, you're not 'evidence-based'. This vagueness is either the result of carelessness bordering on incompetence or malicious in the form of needing 'wiggle room' to disavow failures.

So as usual I'm going to supplement the system with my own inventions to attempt to formalize and science-ize what's missing.

Let's start by defining the 'box' that identifies the ranging, paused, price action that is the first step in trading this breakout system.

I'm going to look at a few different charts. I'm going to start them all at an arbitrary date in the recent past - let's say the start of 2021 and I'll mark that with a thick yellow vertical line. I'm going to use H4 as that seems to be the timeframe the authors use most of the time especially for this method, although they don't come right out and say it, which is frustrating. I'm going to use candles instead of my usual preference, bars, to match the authors' choice.

Here's my pseudo-algo:
1. Whenever a candle prints a pair of lows that are within 5 pips of each other I'll place a horizontal blue dashed line to define the bottom 'zone'. I'll give this zone a tolerance of 10 pips (5 above and 5 below the line) represented by an olive rectangle.
2. Repeat the same for the top 'zone'.
3. I'll demand at least one candle 'interruption' between top and bottom zones. This simply means that consecutive candle touches on one zone don't count as multiple touches. Therefore

This is a 'box' with the requisite minimum 4 contacts to establish a range

But this isn't.

4. I'll use a white arrow (code 343) to show where price breaks out of the 'zone'.
5. I'll put 'LK' on the last kiss candle.
6. I'll mark the buy and sell stops with a green or red solid line.
7. I'll mark the emergency stop loss level with a red horizontal dashed line.
8. I'll mark the profit or loss of a trade with a solid trendline - magenta for a loss and aqua for a win.

So here is the first trade of 2021 on H4 EURUSD with Last Kiss, as I understand it. It is a 70 pip loss that occurs after price breaks out of the top zone, prints a strong bullish candle, that is exceeded the following candle and price returns to the midpoint of the consolidation box. Please let me know if any of this is unclear or if I'm missing important info.

The authors don't mention if we need to wait for a current trade to complete before starting a new one, or if we can re-use bars that were part of a previous zone, so I'm going to select rules that maximize the number of entry opportunities.

The next trade is a 60 pip win, take-profit calculated using a zone that occurred the previous year (2020) so you'll just have to trust me this time.

I continue this way until I've identified all the LK trades. I'll stop spamming my own thread and update you with some tallies later.

But first here's an example of some more 'reading between the lines' problems that crop up because the definitions of the zone are too vague. Is this a valid (losing) trade? It seems that the price drop outside my tolerance zone between touch 2 and 4 should invalidate this trade (which would be good for the system) but in many diagrams in the book N&W allow price to drop and rise far more than this when drawing their boxes.


Re-reading the section on support and resistance zones I find this diagram that mentions touches 'deep into the zone'. But how deep can it go before it's no longer really a 'zone' and more of a 'field'?



Just as crucial, there is no mention of when a 'last kiss' candle expires. If price breaks out of a zone, but then returns to the midpoint of the consolidation box without printing a 'strong' move out of the box, does the last kiss opportunity die when price hits the other side of the box? When it hits the midpoint? Never?? Does it even matter???

EU year-to-date results
415 pips in losses (from 7 trades)
475 pips in wins (from 10 trades)

So, better than completely random, but not much. Earning 60 pips for half a year's trading might float your boat, and it might not. This might bear more investigation but so far it's not really inspiring. Of course there is a whole passel of ways to improve on this, (trailing, break-even) but most of them don't have anything to do with the last kiss method.

One thing I noticed straightaway is how many great trades are not taken because price doesn't return for its 'last kiss'.
It got me wondering what's actually important in this system? It's the break from the consolidated range, not the need to get a better price by waiting for a retrace.

What if I told you (Morpheus voice) there's no such thing as support and resistance? What if you tried simply trading in the direction of price movement?



Template attached if you care to check my work.

I was curious if the results of the last Last Kiss test were trend-biased so I tried it on GBPNZD, a famously ranging pair of late. I wasn't as meticulous this time, and gave the zones wide latitude, but that only seemed to make things worse and it ended up 150 pips down for the year to date. Template attached again if you wish to critique my sloppiness.


Even if this could work for someone else I'm sure it's not something I can use in my own trading so let's just move on to the next pattern - the ominously named BIG SHADOW!!

Big Shadow is not a soulless multinational conglomerate or a soulless devourer of fantasy worlds, it is our next 'Naked Forex' pattern to test!

The big shadow is a two-candlestick reversal pattern. It is basically an engulfing candle pattern that has been given a cool new name. It looks like this:


Here are the rules for this trading 'catalyst' to borrow the author's term for an edge.

And here are some 'nice to haves'.

The authors are typically imprecise about defining the width of the range for this pattern, but really there's not much more we need to know about it. Does it actually foretell anything? Do 'many' traders really do 'extremely well' trading this? Let's find out.

Oh I should also mention what the authors mean by 'room to the left'.

Basically, if you look at the chart and there's a lot of empty space left of the current price that's 'room to the left' and it means that price is at an 'extreme' high or low. Of course it means nothing of the sort, but that's what they are claiming. Here is 'room to the left':


Why are all these examples bearish? In another chapter the book devotes about 20 pages of examples to show a simple 3 candlestick pattern but we are only treated to a few examples of one direction of this trade type. Curious.

Once again I start with EU and begin at the start of the year. Because this pattern is so (relatively) well-defined it's easy to find candidates and I start by marking them with a background rectangle. Then later, I can determine whether they're successful predictors of price direction, and how successful and whether some rules, like the 7 bar 'room to the left' condition is worthwhile.

The mandatory rules I'm following are that the shadow must engulf the prior bar, and I'm including my own condition for determining the minimum range of the shadow, which is that it must be 50% larger than the average range of the past 7 candles. A more relaxed rule is another invention of mine - the requirement for closing price to be within 10 pips of the high or low.

I find about 30 or so signals within the past year on H4. I change the background colour so we can more easily see the rectangles and attempt to mark the 'zones' as well as I can using my poor understanding of what they are. Keep in mind that despite the difficulty, this is still much easier than trying to find them as they're forming in real time. In other words I have the benefit of 'info leakage'.

You can see the distribution of the signals seems fairly random and they don't accumulate at extreme highs or lows necessarily, nor at all turning points.

Now to check how often price hits a 'zone' before price retraces to the opposite end of the signal bar (which is where the stop loss would be placed).

The results are interesting (for a change!)

But first let me reiterate how much I dislike this 'zones' idea in general but especially for taking exits, as I am never sure how ambitious or cautious I'm supposed to be. In my own trading I favour taking predetermined (if not fixed) distance exits for various reasons, but largely because of the avoidance of unwanted, unexpected variability in profits.

What I do like is that we find a definite statistical edge to taking these big shadow trades (without paying attention to zone placement!) The win rate and pip gain is about 2/3, which is what I usually find is the minimum requirement for any non-random result.

Naturally more investigation is required before we can say we've found something worth basing a system around, but it's not too soon to ask why this seems to work. I don't think the engulfing pattern is the most crucial factor. I've tried out tons of engulfing candle systems and found them mostly random. As a lover of momentum I suspect the big range candle is the most important element. A break of this candle is a momentum sign and trading with it is a good, Livermorian practice. The tall bar also produces a wider stop-loss giving the trade a better chance of following through, although possibly at greater risk?

To be sure I would need to check to see how this stacks up against simply trading every tall candle, for example. If the results aren't as good then maybe there is something special about the big shadow pattern, violating my beliefs about pattern-based trading and shattering the foundations of my current trading philosophy. This would be psychologically awkward, at first, for me, but I'd rather be the best trader I can be than stay comfortable with my illusions.

However the results could, again, be trend-biased. This 'catalyst' almost certainly works better in trends, and it loses in ranges. It doesn't seem to matter how much 'room to the left' there is; a sustained range could be deadly.

If you're interested in having me dig deeper into this please let me know otherwise I'll just continue to the next patterns - whammies and moolahs - which is just a silly way of saying 'double tops and bottoms'.

Hi Traders.
I'm building an automated trading system and I'm going to run it on cryptocurrencies. Because I am a newcomer to Forex Factory, I am not allowed to create a forum, so I wanted to know your opinion about the back tests I took here. The back tests are within two years. They are with leverage x 3 and without any money management system. I am happy to hear your comments and suggestions.

Ok, now let's look at double tops and double bottoms. I mean "wammies" and "moolahs". I'm guessing, though it isn't stated, that the authors picked the silly sounding names because Wammie makes a double bottom shaped "W" and Moolah makes a double top shaped "M".

The authors make a number of dubious statements and make no attempt at proof. I am immediately on guard and skeptical of any claims regarding these formations as they have been tested and debunked since nearly the dawn of time, and have no better long-term performance than your average farmer's market clairvoyant. Cycle theorists like W.D. Gann and J.M Hurst have some interesting things to say about the "why" of double tops and bottoms, but that's outside the scope of this post. Recall that Aronson in 'Evidence-Based Technical Analysis' found no evidence of any benefit to his double top and bottom strategies, though they were simple.

But don't despair because Nekritin & Walters have put a twist on an old yarn. These are not regular doubles, but ones with 'optimal trade characteristics'.

Let's look at the 'seven market truths' that must be incorporated into the classical pattern.

Now let's go over this point by point
1. Often? How often? I smell bullshit. X
2. I suggest this is ridiculous. X
3. Exactly how would a market continue to make higher lows without moving upwards? Classic bullshit. X
4. Well, I think we've already seen how these 'zone' touches are less than reliable. X
5. See point 3. X
6. In fact this is true whether it's on a zone or not as my current research is showing. ☑
7. Same here. ☑

Well 2/7 is probably good enough to pass a school course these days (where the lowest grade is 'adequate') but for this old-timer that's a lot of bun and no beef.

Now let's look at the seven characteristics of the 'wammie' pattern.

1. The market touches the support zone twice.
2. The second touch is higher than the first touch.
3. There are at least six candlesticks between touches.
4. The market prints a bullish candlestick on the second touch.
5. The trade is entered with a buy stop a few pips above the bullish candlestick.
6. The stop loss is placed a few pips below the first (lower) touch.
7. The profit target is the next zone above the wammie.

Moolahs of course, are the same but inverted.

Here is a moolah with the first 4 characteristics marked.

So far we can credit the authors with making a well-defined pattern description but they aren't done yet.
That's a lot to keep track of. Recall Aronson once again. The more specifics you provide for characterising a trading signal, the more likely it will be to give the desired result but the less 'cognitive content' it will contain because it becomes less surprising with each new criterion. Adding complexity allows for overfittting.

If we can even find a wammie/moolah that meets all these added conditions, the chances that it will perform as the prototype model performed are slim.

Still, I will do the standard visual backtest and see if I can find any candidates and see how they turn out, provided the sun doesn't burn me to a crisp first.

We have the rules and I'm starting on D1 using the same chart timeframe as the examples in the book. I begin as usual, with EU, starting Jan 1 this year.

I start by drawing the 'zones' in the same rough way that I've done the rest of the tests, using the limited description provided by the authors. I don't use any 2021 data to mark the zones, I use 2017-2020, as we can't know what zones will be occupied by future price in 2021 until they happen.

I end up with something like this which looks fairly regular, fairly consistent, but I'm already concerned we won't get any trades at this timeframe in 2021.

Here are two bars with nearly matching highs. Are these close enough to the zone to count as a first touch? I guess so, I'll mark it #1.

It doesn't hold though so now I'm checking the zone below that. Now I see two touches and they are separated by six bars and the second one is lower than the first.

I mark the SL a few pips above the first (highest) touch and put the entry stop just below the first bearish bar after touch 2. I'm now so close to the target zone that it seems unsporting not to aim further, and in this case it works. 160 pips!

The target actually touches the zone line, but I guess we can't depend on that happening each time. What if it had missed by a few points? The trade would have returned right back to the stop loss. So what is the rule I need to invent for exits? How close to the zone line does it need to get before I can slip in a breakeven stop?

update: I also wonder if I missed a (failed) long entry on this same bit of chart history. I see a rising double bottom but I didn't mark it as the bottom is too close to the next 'zone' barrier. However the authors clearly state that we should seek out trades where the second touch is 'much further from the zone'.

Here's a piece of price history where I can make a case for both a double top and a double bottom. Luckily the double bottom bullish candle is never broken upwards so that outcome is clear and profitable. It almost seems like this event (breaking a candle at one extreme or the other) is more important than the actual zone touches - do you think?
Almost 150 pips - not bad.