Now Reading
Lessons in Price Action – One shot, one kill
48

Lessons in Price Action – One shot, one kill

by ScottMarch 17, 2017

I had planned to make todays post a continuation of the system building, but we have an excellent risk/reward opportunity in Crude today, and there are some subtle clues it is worth investigating.

What story is the tape telling us?

Once upon a time, there was an epic battle, my institutional trader friend Jehan calls it the “Battle of the shales and sheiks”.

bttle

Like we were talking about yesterday, trading ranges end when participants are so exhausted or bored they give up. A useful proxy for “we are in a trading range” for system building is 20 bars or more of inconclusive bullshit.

This is very handy shorthand. 20 bars from the previous trend, the trend has lost influence and trend continuation goes from being a high probability to a coin flip.

Generally we get a couple of clues as to breakout direction. These clues are from interpreting each bar in the context of what we would expect.  The bar I’ve marked as “failure to go up here is bad” is a hammer candle, threatening a breakout. That was a particularly good opportunity to win the day, when breaking out was a very short distance away. Not only was the buying pressure weak, it was so weak it couldn’t even try. Couldn’t even break the candle high. Weak and soft bulls don’t have the energy to land the knockout blow.

We should start to get interested about this point.

What next? The smart play is to buy the bottom of the trading range, because it is a low risk idea. A very small risk for a larger gain. And the three bars off the lows are frankly pathetic. Keep in mind that these are the bars (first few bars off the lows) that will ALWAYS have the strongest and most sensible bulls buying. Look at the previous downside breakouts, bulls reliably stepped in to push it hard off the lows.

Look at this bounce. Not a chance. At this point, we have the odds all lined up in our favour at the time to bet is at hand.

So how do we play things when we see price action like this? We look for a point which confirms our view, and get in when price crosses that line. For trend following setups it is (mostly) better to let price confirm by moving in your direction first. Here I got short on a break of the lows, with a stop at the candle high. You can see I banked half at 2R and half at 4R.

crude

But the past is the past. We banked and yanked and now it’s time to consider a fresh shorting opportunity.

What are the best trades? In downtrends you should be shorting rallies that fail.

Here we have had some clues that the bulls were going to mount a rally. If that rally gains legs, we probably see a short squeeze. If it fails, the rally has failed and the bulls who bought in the last 4 days will be forced to sell, driving the market lower.

2017-03-17_17-14-46

Today is the critical day. We have a failed attempt to drive the market up. Right now the bulls can still make a tentative case, but if the bears win here, the bull argument will literally sound silly.

These are the points we wait for as traders. The point of inflection (Mole’s term) where one side or the other is forced to admit defeat. Go short if price breaks the daily low, with a stop just above the high. 50% odds of a big win. Betting odds.

As far as the indexes go. I remain convinced the rally stinks and will likely fail before the old highs, but we get more information today. If the bulls are strong enough to break out, then we really should have strong buying pressure off the open today. Otherwise the door is open for the bears, just a little. We certainly don’t have betting odds yet so this is all rather interesting but keep your powder dry.

2017-03-17_17-22-08


About The Author
Scott
  • ZigZag

    Good stuff. Question, how do you account for contract rolls?

  • Grant

    Scott, it needs to be said that outside period bars are key here. The 3/14 op hammer candle will likely be a capitulation / exhaustion candle for longs and possibly for shorts. If the longs can break the shooting star then they have something but the likely bet is to get short at the break of the ss. I have also found that a significant move such as the one on CL that was signaled by the 3/7 OP/SS can also end in a trading range started by an OP. This may be the beginning of a trading range that will not end today… I would not be long at the high of the 3/16 SS after the price action to the downside. If no additional OP emerges on this TF then wait for the retest of the 3/16 SS. The outside period is the proverbial “rack” that breaks the back of longs and or shorts. It’s an inflection point.

  • saltwaterdog

    Hi Grant… same page for the most part, but all about odds. The short play on the break of the shooting star is the favorite and to me the long on the break of the SS high is the underdog, or the money line bet. Lower odds = higher payout imo. It is possible that the SS was a retest of the 3/16 low, no?

    Always trying to canvas the other side but in the end, I agree with you for the reason that the daily RSI made its lowest low at the price low. This is how I use RSI — typically an RSI low <40 that is the lowest low does not represent the final price low…

  • AcoBrasil

    Another education post for me, and opportunity to improve my buying strategy.
    I instinctively tend to enter at the range bottom or following a downside break of the trading range without waiting for a confirmed reversal move. If I were trading this, I would probably have bought near the bottom of the range ($52) the first red candle ($50) or when I saw the hammer move (around $48.50). Probably confirmation bias since the range has reinforced the perception that the $52 range is a safe bet, and that anything below will soon correct upward back to the range.

  • AcoBrasil

    Good question. The contango is greater on the short end that I expected to see.

    https://www.investing.com/commodities/crude-oil-contracts

  • Grant

    Yes on the RTV, but after the price move to the downside, I would be skipping the first long setup. If anything, I would be long at the high of the OP bar anticipating a reversal lower at the low of the SS or more likely a trading range. I would not be taking trades after today’s bar if we get an IP. I would be waiting for the next OP.

  • saltwaterdog

    interesting… so if we get a SS yesterday followed by an IP today, which arguably is a RT of the SS high, you don’t take that sell setup?

  • Grant

    Personally I don’t. Here’s why. If after the 3/16 OP the price action does not resolve lower after 2 bars or retrace quickly to the long side after 2 bars then we have a trading range. The range will only be resolved after another OP bar but the OP may not materialize on my chosen TF. I may have to look at a shorter TF but it will be there. Fractals and all that….

  • saltwaterdog

    very interesting take on the OP as a mile marker for change in market type. Certainly shakes everyone off the bus which creates the condition for a chase (“let me in!”)…. I’m trying to quantify any correlation between size of a run vs the range that may follow. it seems to me the larger a win I take, the more consecutive losers I’ll also harvest if I take every setup after. CL had an 11R winner on my TF followed by 11 losers on the next 12 entries, 9 of them for a max 1R loss.

  • Grant

    Spot on…. hence the OP friend vs foe…. It’s not always what we want and may be a harbinger of a constraining trading range but at some TF will lead the way out of the trading range, if that makes any sense, lol.

  • Yoda

    Tape is starting to look bullish again

  • BobbyLow

    It’s funny how different lenses can see either the same thing or see something completely different while looking at the same trading vehicle.

    Regardless, I look for trends and there is still no trend on /CL. Scotts entry at the break of the March 7th Low was excellent. March 8th provided a $2.59 Dump followed by another $1.63 on March 9th and 10th. By this time, I was thinking Oh Shit, after waiting 2 months for a new trend to begin, I can’t believe I missed this one but decided not to chase it. Then over the past 5 days we’ve had 3 Long Tailed Doji’s and 2 other small candles with price drifting where? Yep, upward and as of this moment, price is within $1.02 of the March 8th close.

    So FWIW, as much as I miss trading Crude, I’m glad to have at least temporarily taken a break from it. Part of me wonders if crude has become a more or less a permanent range bound trade but Scott mentioned the “Battle of the shales and sheiks” and there should be a winner of this battle. My main question is how long does this fight last? It’s like back in the old days saying something like “this rally can’t last forever”. And the answer is no it can’t last forever but it can surely shit all over your account if you keep trading against it until it ends. So perhaps there is an out and out victor between the Shales and Sheiks and maybe it’s already over and maybe it ends next week. or next month or . . .

    This week has been relatively quiet for me with what appears to be 2 Break Even trades. (I will close out my Long NUGT by the COB today {Don’t carry Miners over the weekends} and I’ve already closed my Short Nat Gas and flipped it to a Long.) I’ve changed my method on Nutty Gas by including A/H on my hourly chart and will carry it over weekends now. I’m also carrying a +.5R position in Long Silver.

  • Scott Phillips

    Grant, Nice to hear from you mate :-) Outside periods are critical clues towards the end of trends, while falling short of being “key”. The one at 3/14 is a classic.

    In trading ranges they are normal and to be expected.

    I’ll cover what an outside period does and doesn’t mean in a future post, but like a lot of Ivan’s work it’s not a good edge either way. All it means is “longs got faked out of their positions, so did shorts”.

    The key is context. If it’s making an outside period while it should be trending, that’s important. If it is in a downtrend, tries and fails to go up and closes outside period down – thats consistent with a downtrend. If it’s making an outside period in a trading range, that’s entirely normal.

    This is typical of Ivan’s setups, which work extremely well for certain types of tape, but overall are either no edge or a tiny edge.

  • Scott Phillips

    An inside period followed by a shooting star is a strong edge in trending tape. The other way is more complicated and resistant to simple answers

  • Yoda

    mmm, perhaps not

  • Scott Phillips

    You have a misconception. A range doesn’t resolve with an outside period. It resolves with a close above or below the range.

    To take a random example, the chart open on my screen right now is SB, which I traded to the downside the last few weeks from the shooting star break. Not an outside period in the range, though an outside period signalled the end of the trend.

    https://snag.gy/IUSVeY.jpg

    I don’t know why you think there is some requirement for a particular shaped candle to be part of any price action, but that’s straight up wrong :-)

  • Scott Phillips

    Excellent question. I look at both months for clues. I look for setups off the current month and if it’s close to rollover take them on the next month.

    Another valid option (better for longer term trading) is to use correctly back adjusted data, but there is no “correct” way to do this and you need a consistent methodology. I like esignal for this, since you can change the back adjusting methodology in cool ways.

  • Scott Phillips

    I’m of the firm opinion that systematic trading is far superior for all but the most highly skilled of traders.

    In fact, most of the best traders I know cannot do anything like this kind of tape analysis, and I know many who can who are, shall we say, broke and fake.

    What you can use it for, aside from making interesting blog posts, is identifying specific conditions where you might want to build a trading system around. My mean reversion system is built around retests. My trend following system is built around hammer candles in an uptrend.

    You could build a very good system about buying the lows of trading ranges, for example. The candles themselves, aren’t as critical as the actual level of the market.

    What will NOT work is saying “this specific candle pattern is an edge” and trading it throughout the tape. Trending and rangebound markets and low and high volatility markets are so different it is literally impossible for something to work in all seasons.

  • Scott Phillips

    Well the market failed to break the low of the setup candle, so no entry for me :)

    This is quite suspicious in the context of AMPLE bearish evidence. It reduces the odds of a bearish conclusion, and increases the odds of a mere retest and short squeeze.

    I’m saying you could be right here :) But as a general rule you can expect 10 bars/2 legs minimum after a long running range. If the second leg down is bigger than the first, likely you have a major sea change. If the second one is shorter than the first, thats also key information.

    For example I was short Sugar today, and even though I got short on the downside break, we have clear rejection of the bear case. That, in context is not bearish at all.

    https://snag.gy/IUSVeY.jpg

    Bobby, your system is very good, and I’d urge you to stick with it in the long term

  • Scott Phillips

    Contango driven trends tend to last for a long time, as professionals are comfortable holding them while amateurs are comfortable fading them.

    My friend Andreas Clenow puts great stock in a bearish contango trend

  • BobbyLow

    Thanks Scott. I’ve had a great teacher. :)

    I’m actually experiencing serenity while trading some pretty volatile stuff so either what I’m feeling is real or I’ve gone completely nuts. LOL

    But seriously, it still comes down to what you’ve said in many different ways. If you think you might have an edge (that fits your personality), back test it with what your rules are for entry, and what your rules are that tells you when to close your trade. Then forward test it and tweak it if necessary. Then execute what your rules have told you to do. :)

  • Yoda

    Wow! @ the close of session

  • Scott Phillips

    Not easy but simple my friend :-)

    Complexity is laziness, the best solutions are elegant, simple and robust.

  • Scott Phillips

    Until we break out of the range one way or the other, the tape doesn’t mean much.

    The highest probability is a downside break with limited downside, followed by another leg up.

  • Scott Phillips

    There is no correlation, and no way to tell which trends are going to be the outliers that make your month/year.

    You take them all, and build a system that takes into account the inherent fat right tail distribution.

  • Mark Shinnick

    Yeah, went long volatility damm near the lows and holding.

  • Yoda

    Thanks. I like that plan.

  • Grant

    Scott, there are many examples of what one would call a breakout only to move back into the trading range (consolidation) only to churn some more. IMO, trading an OP from the candle b4 the OP is a great way to start a campaign as trading the back side of an OP is an edge as long as good campaign management is used.

  • http://greenlander1.blogspot.com/ Greenlander

    Unloaded short term oil plays and sitting on 1/2 of long term in APA/ XOM. Looking for more oil wkness to put the other half into play via XLE/ XOP.

    Bought a lot of SNAP 25 July calls.

  • http://www.captainboom.com/ captainboom

    This is a new idea for me. Thanks.

  • Scott Phillips

    I hate to break it to you, but there is a fundamental problem with your approach rendering it completely flawed, and completely unusable.

    Suggest you reread last weeks posts. The problem is that your hypothesis is that after an outside period is a good time to start trading a system without an edge (yes, the system you turn on is neither a positive nor a negative, edge, it just chops around). If you test that hypothesis through back testing you have a literal random result. Not only are you testing the interaction of many different setups working in completely different ways, but also the stop distance and profit taking mechanism, and also the original hypothesis that outside periods are somehow magical.

    This is why you made 25R very quickly when you turned it on, and gave it all back very quickly also. Polishing the turd by employing equity curve trickery on it cannot save it. At the end of the day you either have an edge or you don’t.

    Sadly, this approach is not going to work. I know its a bitter pill to swallow, and you’ve done an enormous amount of diligent work towards being a pro trader. Much respect for doing that.

    Suggest you start again with something that’s actually an edge, and able to be proven to be an edge. We don’t need to find esoteric knowledge or obscure theories. None of the professionals I have spoken to have ever based working systems on anything even remotely controversial. Trends, ranges, vola breakouts, momentum, mean reversion… these are the low hanging fruits.

  • Scott Phillips

    Further to this I should tell some (very vague) details about Bobby’s System. He is trading a shorter term trend following system with wider than usual initial stops.

    This is an approach which is likely to be a rich vein to mine for anybody developing systems of their own. Bobby has not sought to reinvent the wheel or come up with an overly complicated Rube Goldberg machine.

    He waits for a trend, enters with a decent stop, and exits when the trend looks over. Simple hypothesis, easy to test.

    Complexity is laziness, and with trading systems simplicity is the ultimate sophistication.

    Compare and contrast with legendary trader Ed Seykota’s system.

    If EMA 140 > EMA 80 take short trades only
    If EMA 80 > EMA 140 take long trades only

    If price closes above a previous 50 day high, buy
    If price closes below a previous 50 day low, sell short

    Trade management: For long trades, if price makes a new 50 day low on a closing basis, exit position. Vice versa for short trades.

    3 rules, and billions made.

    What’s the lesson? Every extra rule you add makes things more fragile. Simple systems are robust. Complex systems fall apart at the most inopportune times.

    The worst mistakes I have made in my system building have been the temptations to add extra rules.

    The worst thing about non-robust systems is you produce a 5 year backtest showing a max drawdown of 15% and then it is entirely possible and normal to get a 40% real world drawdown. You don’t understand how or why, when it never showed up in your simulations. You just got fooled by randomness and paid the price for too much complexity.

  • Scott Phillips

    It’s a nice guideline, that you can expect choppy tape in trading ranges. In a trading range if you actually start to get decent consecutive runs of one sided price action, that is a sign the range is about to over.

    Literally the opposite of trends. Which is why “one size fits all” systems rarely work without a filter of some kind

  • Scott Phillips

    Don’t get too attached. Its wrong.

  • http://www.captainboom.com/ captainboom

    The idea is using a bar before a signal bar. I wouldn’t use it without testing it.

  • Grant

    Entry is best at the inception of an OP bar, not after.

  • sutluc

    Those 48.29 buys were the last of your short coming off?

  • ZigZag

    Upon reading this all again after market and after some St. Patty’s day libations, so glad we asked for both Tape Reading and System Building stuff. Win Win for everybody!

  • Scott Phillips

    Yep. But it’s another case of lucky not good, my rules are optimised for a decent result in trending tape, so they make me look better than I am when the market is matching up with the system.

    The failure to go down on Friday with strong bearish evidence opens the door for a bullish case. I am forced to now consider a long in Crude.

    Flipping direction is emotionally tricky for most people

  • Scott Phillips

    I’ll get back on the System building next time. I have a lot of ground to cover and I keep getting sidetracked.

  • Scott Phillips

    Test everything please. Especially my stuff. I’d be very disappointed if you guys took stuff at face value and traded it without deep testing. I’ve been wrong, and very wrong, before.

  • ZigZag

    Sidetracked. Distracting. I think that my favorite quote of all time is Jack Schwager’s. “There’s a million ways to make money in the market. They’re all just really hard to find.”

  • Scott Phillips

    I re-read his books once a year over Christmas time for inspiration.

  • Trouzzer_Snake

    Re: CL – It seems to me that if Fri’s IP is breached low then that would trigger a bearish set-up too good to be true. However, being the pessimistic SOB that I am it makes me think it IS too good to be true.

    Vol was light on Fri which seems odd if this really is a line in the sand for the Bulls. That would mean Bulls have either already capitulated (which would have made it easy for Bears to take this down further, which they didn’t) or that the bears didn’t show up for work. Both options stink.

    My theory is that this is a bear trap waiting to be sprung.

    Any thoughts?

  • Scott Phillips

    My take on it is that the downside of the shooting star *should* have been broken. That it was not lowers the odds of a bearish resolution, though the odds are still with the bears.

    Volume doesn’t mean anything – forget about that :)

    It doesn’t change the odds to be in the bulls favour yet.

  • Trouzzer_Snake

    Are you looking at the IP breach to provide evidence one way or the another?

  • Yoda

    A good filter can improve your win rate, but I think it is like adding another variable into one’s equation. Could lead to a blurred vision if wrongly selected. The zero is what I would call a good filter, but there are others. I’d say it’s what Mole would call Context, which he is damn good at reading correctly. I’d love to see a picture of the daily zero if it still exists.

  • Scott Phillips

    NEW POST – And todays has some really useful stuff