Questions for real pros: How to trade with reference points only?

Over the past few years, I've been hanging out here at FF. I've particularly analyzed two systems (or theories of systems), collected statistics, researched how prices behave under certain situations, and so on. Despite conducting sometimes intensive research, I haven't been able to consistently make profits. As a result, I've taken a step back and considered starting anew.

Throughout the years, I've consistently observed that the way prices move is not significantly different from random walks. For example, the expected profit is strongly correlated with the holding time/trading duration. This apparent randomness of the market also contributes to why it's challenging for me (and probably the other 95% of people hanging out here at FF) to consistently achieve profits: The risk-reward ratio seems to always be inversely correlated with the probability of winning.

Furthermore, over the years, I haven't found any significant added value in indicators (except for smoothing out price trends). Therefore, I want to go back to the basics of trading and solely focus on reference points (such as prices of an asset from 0, X, Y, and Z minutes ago or the OHLC of a higher rolling candle) in my learning process.

If the early Japanese rice traders and tape readers were able to consistently make profits using only key price levels, it leads me to one conclusion: trading with reference prices alone must be possible!

As I personally want to grow through this question and thread, I kindly ask the pro-traders among you who would consider themselves profitable and who master trading using few reference levels (not necessarily relying on charts themselves) to help me out! Thank you in advance!

Don't worry, I'm not expecting one-on-one mentoring here. However, I'm particularly interested in two things: the logic and theory

1. behind entries and exits when making trading decisions solely based on reference prices,
2. and how one can escape the trade-off between risk-reward (RR) and the probability of winning. For example, how can one expect a higher probability of winning than 0.5 with a 1:1 RR ratio?

Thanks a lot in advance for taking the time and drop some hints here and there! Have a nice discussion everyone!

Here I want to start the discussion with some general ideas I had/found on the reference line trading concept:

As mentioned in the 1st post, I imagined two ways of trading reference lines so far (of course, you can use any line you want): 1. prices at arbitrary times back in time:
Example from CrucialPoint (unlike CP I'd like to think about drawing these open lines with every new minutes/tick, i. e. I think one shouldn't restrict themself to just use the pre-set time frames, but start a new time frame every tick/second/minute:

1. For reasons of simplicity, let's say we're only looking at the prices at the very moment (P0), 15 minutes ago (P15), 1 hour (P60) ago and 4 hours ago (P240). Since we only have 4 prices available that are associated with a reference time, we can classify the trend by looking at the order of the 4 prices. In the best case we'll find the prices sorted so that P0, P15, P60 and P240 build an increasing or decreasing order.
2. Let's say P240 < P60 < P15 < P0. This would be a consistent up trend. However, for just 4 key prices there are already 24 distinguishable positional combinations (orders in the sequences of 4):
   
   Attached Image (click to enlarge)

   
   
   Which of these combinations would you trade, which would be too "ranging" for you? Why?
3. However, the order seems to be just a qualitative measure, a structural indication of the market. You can have a weak or strong trend with a low and high return as well as a small or big range (depending on the volatility of your asset during that trend or range). Thus, I guess looking only at the market/price level structure might be not enough. Thus, I should maybe look at the difference in the price levels between P0 and every other prices, P15, P60 and P240.
4. Is it neccessary to consider all 4 distances? Do the other distances (i.e. P15 - P60, P15 - P240, P60 - P240) matter, too? Cannot I just read these differences from the 3 distances with P0?
5. What should I do with these differences. I obviously have to classify them to know which price change (either positive or negative) is weak or strong. But what to do then with these difference classes?
6. How can I use these difference classes to build an entry decision tree around it?

Alternative: 2. Calculate a rolling higher candlestick (e.g. a 60 minute candlestick) with every new tick or every new M1 candle
Example from 030985:

 

1. In that scenario the only way to measure the trend strength would be the difference between the close (P0) and the open (P1) of the candlestick. Since I could only use these two key price levels, that I'd be not able to investigate the market's (trend/range) structure.
2. However, in comparison to strategy 1., I'd have two further important measures, the high (Ph) as well as the low (Pl). Assume I'm in a short trade, then I could easily measure the counter-trade force as difference between the close (P0) and the low (Pl) of the rolling candlestick. If this difference gets too big, I could just exit the trade.
3. Would you say that in case of a short trade, the high would have any use?

Maybe just using 1. or 2. alone is not the best idea. Maybe one could use the 4 key line method (alternative 1.) and just add the floating high and low (over the last 4 hours (in case P240 is the price that lies most far away in the past)?!

Now it's your turn! You are free to either comment on my thoughts or (or course) add your own ideas that we can discuss later on. Please be polite to each other and accept the opinion of each other. Every participant should accept that many people can have many different points of views and can have a multitude of perceptions what consist their "truth" about the market.

I think your over complicating things. If you have time lets get on zoom. Its way easier to explain than typing in a thread

Just out of interest... Why are you inviting everybody to zoom?

Just to help anyone with limited understanding of how the markets work. I see many blowing accounts and jumping from one strategy to another trying to find the Holy Grail. The closest your gonna get to the Holy G is Risk Management. Im no Guru just a regular guy who has been trading for over 30yrs, and yes i get Loadsa trades wrong but i know when to get out and not add to a loser as i did many times in the beginning. I just like to help if and when i have time. Im not saying my system is the best, it isnt. there are thousands of ways to trade and no one size fits all. Its a Traders job to find his own path. But if you join me on zoom and practice what i show you it should put you miles ahead of where you are now, And the best part is i do for free.

It takes Years and Years to become a Lawyer, Dentist, Doctor etc. So you need to be willing to put in the hours of work to get anywhere in this game. And when i say hours i mean at minimum 20hrs per week to start with

You can join us on zoom at 9pm UK time or 2hrs 30mins from now just send me dm and i will send you link along with AntiCre 2mins before we start

If I'm ahead of you? Will it put me back? Just kidding. Happy zooming with AnCre

is that too late for you, not sure what time zone your in

https://files.slack.com/files-pri/T0...9/eurusdh4.png
Do you see similarity?
Regarding rice trading: It has really fine edge if someone understands how rice works.
Same as FX

https://files.slack.com/files-pri/T0...W/eurusdh4.png
This is the right ling

I have to agree with T721.....

dont forget the japanese rice traders didnt use indicators apart from their candlesticks and possibly trendlines.... the only "indicators" they must have watched is rain and drought during growing season ...and definitely not bollinger bands or stochastics or all those other useless ones that people watch instead of candles

and also dont forget the old timers in early 1900's mostly viewed daily charts they drew by hand at home on graph paper...the big ones subscribed to chart services. they didnt trade the 1 minute or 10 second charts like the goofballs do now....the only way you could scalp intraday was if you owned a seat or membership to the exchange...watch some old youtube videos to see how they did it LOL

I only use price action and volume with support resistence...you dont need anything else especially not that indicator you posted which doesnt seem to be of any use honestly.

so my advice would be to develope a simple system with as few components as possible

says to sign in

Nope, says imageI don't need your sign in

Oh, you mean to view the image to sign in. I will find another way

You're, not your.

You're, not your.

make your Analysis simple

Alpha decays over time. If anything, if it used to work in the distant past you should not expect it to work in the future. With ML and ARIMA models picking up on patterns far better than any human, finding success using price alone is not a reasonable expectation.
That’s a good point. In other words, alternative data (or anecdotal evidence) could have been attributed to success even then.
Simple systems may work best for discretionary traders, but not so with algorithmic. There’s a white paper called the Virtue of Complexity in Return Prediction by AQR that shows that adding additional components into models does add value, and the best models are incredibly complex.

https://papers.ssrn.com/sol3/papers....act_id=3984925

Would you be please a bit more precise? I'm only looking at 4 data points (or maybe a 5th, my entry price). How more can I reduce it further? Where am I overcomplicating it?

I previously tried ARIMAs, but I've only seen little to no advantages. As you said, for me they projected too far into the future and were too less adaptive to profit in the long run from a rigid prediction.

Thanks a lot - will have a read. Cheers!

Here's something else to show the lines towards I try thinking. All with the hope to find someone who's willed to discuss my questions from above. Of course I know, that most here will think that I'm completely nuts and that they trade successfully with their more simple reasoning and methods. Yet, I never found someone who could resonably answer the questions above and could show me the conundrum of trading profitable, i.e. essentially breaking the winning probability-RR conundrum. (I guess most never did the stats to find that conundrum, and thus never thought about why they are profitable in the long run - what's it that makes you profitable). This thread is not a thread to become successful (although this would be nice) but to finally understand how it is possible at all to be successful on a chart that behave like a random walk in almost all aspects.

I took thousands of candlesticks from a higher TF. To investigate the momentum impact that some say would exist, I entered long after X pips has reached from the open line, aiming a 2X profit. This is what distributions I obtained when only considering up bars, down bars, or when I don't differentiate between the bars at all. The 1st row shows the results without having a SL, i.e. before hitting 2X after X has been reached, the bar could have expanded it's current low much more to the down side. Aiming only for a very small pips profit, I'd be almost guaranteed to make 1, 2, or 3 pips regardless of the color that the candlestick will finally end. However, there are two down sides: 1. The DD is often really huge. 2. The very rare cases when I miss my profit target, the candle would end in a very deep red number, and thus in negative pips that eats up all small profits that I'd have made 99 % of the time. Add trading costs and it is an overall loss.

The 2nd row shows the distributions if I'd use a SL of X, i.e. once price reaches +X pips from the open, I'd have exited if price would have retraced to the open before it reached 2X pips.
And here some magic seems to happen. If I'd know in advance that the bar would end bullish (i.e. the close of the matured candlestick could be just 1 tick above the open), the probability to make up to > 40 pips profit is higher than 0.5 with a SL (this would give me an edge on at max approx. 58 %). I wonder whether I can find an indication (price path wise) whether the odds are towards a bullish or a bearish bar at the entry criteria?!

And no, the most obvious idea to wait until a lower wick of the average lower wick size has been formed, won't increase the likelihood.

I still have a few ideas to solve this riddle myself. If I then won't find someone to explain to me how to solve it, I'll either really go nuts or I'll give upor both.

As soon as I read about using a "reference point", I immediately thought "oh boy, someone is going to mention CrucialPoint"... Well, it didn't take long.
IMO, you're on the wrong path if you're trying to find some magic bullet with a reference point idea. Especially if that reference point is out of context, i.e. a rolling 1s, 1M, 5M, 15M... reference point. There is no point in doing statistical studies with such a general approach, there is zero edge here (Adam Grimes has done some work to that extent).
I know you never mentioned trying to find a magic bullet. But actually, you are.

A few comments on some of your points:
- Regarding trade-off between RR & WR: Yes, there's nothing new here, and it's not really a conundrum as you put it. It's also why everyone knows that when a guy shows up and says he's finally found his system, with a 2.5RR and 63%WR you automatically know he's bullshitting, without the need to dig deeper. On a general sense, this trade-off can be a starting point on what kind of method fits you: catching big winners while getting punched in the face many times with reversals, fake-outs, etc... Or, getting more consistency but with small wins. I don't think there's a better approach, but I do think there's a better approach for a particular trader, i.e. the whole "fitting your personality" thing. (from a pure track record perspective, the first approach seems much more profitable in the long run, but again not for everyone).

- From my experience, trading can be as simple or as complicated as you want it to. There is no correlation with complexity with profitability. But, fitting the personality (& inner beliefs) is important. Complexity is actually fine, if it's what you believe is needed to get the upper hand vs the rest of the market. There are more than 250 episodes on Chat With Traders. If you listen to the majority of them, you realize that, at the end of the day, what approach / strategy you have to the markets, is in itself insignificant. That's an important realization.

- From my perspective, one of the most underrated aspects of trading, is market selection. There's this whole "trade one instrument until you know it by heart", maybe there's some validity to it, but I've yet come across a seasoned trader, who traded only 1 instrument, and when shown 2 charts was able to define with confidence which one was his preferred instrument, and which one was the random. Makes you think about "knowing how an instrument moves". Personally, I have no clue how a particular instrument moves. But I do know, that between the 2 instruments below, I know which one I'd rather trade. (no cherry picking, just current charts).





Spoiler alert: on the second chart, you can stick any moving average method, any macd, rsi, stochastic, fib retracement, pure price action, fundamental approach, volume price analysis, value area trading, trendline trading, naked trading... it would be hard not to make money trading long continuations.
Give me the first chart and I might as well flip a coin for any trading decision.
That's also where your RR/WR conundrum might be solved: given certain market conditions, there are higher probabilities of out-sized returns in a certain direction.

Hi Anticre, you will go crackers unless you start spending time separating things out. From what I can see, you're simply looking for an edge across MTF but you're wasting time in not understanding what it is you're specifically looking for or what's gone before you.

Reference prices; CME traders for decades worked off pivots (OHLC)
Tape reading: requires volume, is not the same as OHLC reference point.
Honma <> standard rice traders either now or in 18th century.
CP <> any of the above either though there's some crossover.

Quant modelling is its own discipline but if you're trying to brute force your way through to discovering how say Honma traded you're using the wrong tools.

All CP is saying within that post is that using the open rather than the range and time to determine the strength of the cross above or below open. It's the basis of simple and complex strategies world over, so it is useful but by itself confers no edge.

If you're really trying to unravel anything, figure out what it is you're trying to unravel, and is what you're looking for public or private.

The former is costly in psychological resilience to trade, the latter in your personal time. One way or another, we all pay our dues.