What Works in Trading & Why: Part 2

What Works in Trading & Why: Part 2
The Nature of the Beast


This thread is my personal view about what works in the markets and and why. I should have called it "What Works in Trading and Why for Me" because it's not the only way to exploit market inefficiences. My initial thread on this topic (referring to Part 1) stirred up some ire and controversy. Next to each thread listing it says who started the thread: "NoiseTrader" -- so if my comments are offensive (for whatever reason) then stop now, read no further...

I believe my particular world view of markets is pretty accurate and a good stretch closer to reality than it was 30 years ago. I also don't believe it is 100% accurate.

The Nature of the Beast

The real nature of that beast we call markets, is pretty important because to not see things as they really are in a market exacts a price, and for some a heavy one. So to minimize risk (and maximize reward) I'm constantly working to get closer to that reality and stay in a frame of mind that maintains objectivity -- which is another reason I'm not going to engage in fruitless debates about the merit or veracity of what I propose here. I want to stay in a "formless" (ego-less) state of mind about what I see in markets and defending a specific point of view has a insidiuous way of fueling attachment and undermining what I believe is a more optimal state.

Point 1: Quants and economists generally overestimate the degree of market efficiency. I believe this is true because profitable trading entities (who depend on inefficiencies in markets) seem to exist for more than a season. Traders, on the other hand, especially technical-based traders, vastly overestimate the degree of inefficiency. I believe this is true, otherwise win-ratios would be much higher and trading much easier than it is.

In breif, the beast is mostly efficient, but transiently inefficient, which means profits are possible, but often fleeting (sound familiar?) and that has tactical implications which I get into later.

Point 2: Fischer Black, co-inventor of the Black-Scholes Option pricing model had a pretty accurate world view of market because, his take,on the nature of the beast, manifest in his model has proven accurate enough over time that option market-makers have become by and large pretty wealthy using it. Now, Black said: "Noise makes profits possible."

Somebody said, I should call myself "Anti-Noisetrader", not NoiseTrader because I claim to hunt noise traders in order to detect profitable opportunities. He's right, I am hunting noise and trying to avoid becoming a noise trader myself; but when I (or anyone else) trade on what I think is information, when in reality that information is just noise, then I am, in that instant, a noisetrader and the guy who is fading me probably has the real edge.

The bottomline is this, if you can uncover where noise is occuring in a specific market, you are close to uncovering where profits are possible.

It would be disingenuous to pretend that the analysis I'm going to narrate in the following two trade examples provide sufficient information to make these trades. As mentioned in the Part 1 thread, I use a raft of proprietary stuff to select a group of trading candidiates and I'm not going to reveal specific algortihms -- that's counterproductive -- so my analysis here is accurate, but also partial. These are trades I posted live to a Yahoo conference, that I also executed myself, where I took partial profits, and where I hold core long positions as of 4/15/06. I am going to describe the process fully and how I think about these trades and I believe there is value in going through that process with you if you are of a receptive mind.

Recall, in Part I, I mentioned some paradigms I called "big ideas". I tend to weigh ideas based on two criteria: what works in the market (positive expectancy) and what works the longest (robustness). The markets have already forced me to part with so many pre-concieved and dear notions about what works in markets, that I've sort of let it all go; and I don't care whether an idea comes from a specific school of thought: fundamentalists, quants, technicians, quantum physicist, or even poker-players. If it works in a market and it has worked for a relatively long time then that's meaningful to me.

I said one of the "big ideas" for me is this notion that “Cheap assets outperform expensive assets more than they should” and “Momentum portfolios work better than they should.” The key that allowed me to port the first half of this thought pattern into the moreo time-sensitive world of technical analysis is this: "the short-term volatility of price is greater than the short-term volatility of value." This provides a context for uncovering noise and possible profits in markets, because value is largely unobservable and hard to pin down.

I have uploaded an annotated chart set. You will need to download that and unzip the file and pull the charts up to follow the narrative.

Getting Long Boeing (BA)

Chart 1 is a weekly chart of Boeing (BA). The red and green lines represent a calculation moving forward in time where value buyers are willing to hit the offers of momentum sellers. The logic is this: If I can measure rising equilibrium (value) levels and project where equilibrium should be in a future period, then I can detect undervaluation in the current time period. Restated, if I can buy next month's projected value (which is set to rise) right now, then price is currently below value. The red and green bid lines represent that notion and I post a set of charts like this for subscribers weekly.

At A, B and C momentum sellers had a variant percepiton versus value buyers. The only reason to sell BA at A, B and C is because you think price is going to go even lower. So the expectation of those sellers was at a variance to the expectation of buyers. In short, whatever information motivated the sellers to offer BA was not information, but was, in fact, just noise; and those sellers were acting on noise as if it were information and therefore they were the noise traders -- and you can see that hunting those noise traders has been a profitable thing to do of late.

As price retraced to the bid line again in late January 2006, I was alert that noise traders were present again because at that price level there was another variant perception about BA.

I also said that of the "big ideas" for me is Hart's 7th Law: "That dramatic price movements tend to unfold from price structures that minimize profitable participation.” This notion is much easier to translate into price action, than the last one. What minimizes profitable participation in a market is usually either excess volatility in the absence of market structure or a violation of obvious structure. I'll explain the former now and the latter in the Gold example to follow.

In Chart 2, which is a closer look at the weekly, notice the weekly ranges are failry uniform between A to C. Near the high at B we saw some range expansion indicating some noise trader activity and an intermediate top soon formed. Then BA rolled over and retraced down into the bid line at D. Notice at D, that the weekly range (volatility) expands as that range is greater than the previous nine weeks.

In the higger frequency Daily Chart 3, notice the well-offered selling climax from A down into the bid line at B. This price action is minimizing profitable participation on the long side due to an increase in volatility in the absence of structure, meaning there is no clear thrust here through a high visibility swing bottom, under which stops normally nest.

Steidlmayer's notion of readibility vs. reliability is another idea that's hard to pin down and difficult to codify. I know what Steidlmayer is saying and I know he's right, but it a hard thing to wrap your mind around. We might say that the most reliable price or trade location is, of course the exact low at B. After all, anyone who takes the offer there endures zero heat on the trade, but there is so little to read in the price action at the exact low as it occurs. After the price rejection and reversal day at B, we have more information (more readibility) but we have to pay for that information in the form of a higher price.

The price action that occurred in this region showed up on my radar as potential noise trader activity for all the reasons cited here and more; and I stalked the price action for a few days, waited until the bids started to come in and told everyone in my net conference I was buying it and bought it on 1/26/05. We're still holding core longs as of 4/16/06, but the recent increase in volatility (new noise trader buying) may usher in new structure (a congestion or a new reaction.)

In the post that follows, I'll parse through the logic for Buying Gold in Feb and Mar 2006.

If you want my list of the exact actions we're now taking in stocks (and commodities) and my daily trading plan drop me an email at my blog:

Use the "Send me your BUY LIST" button, right-hand column at: http://www.street-noise.net/articles/

 

Thank Noise Trader, interesting lessson. I have to imagine that your pink and yellow dots on the line mean something. Could you elaborate if they aren't "proprietary"? I am guessing the open and closes on bar charts? I don't have much experience with bar charts.

 

Yes

 

hello NT

I know you have been influenced by the works of others in your career in the markets ... we all are

has the recent publication to a wider audience
of the work of the ny oil trader MF and his "Logical
Trader" methodology had any affect on how you view the markets ?

( if this is not the place for me to ask such a question then I will PM you ... and apologize )

 

NT,

When you reference Steidlmayer's methods, are you referring to market profile? I thought I read on the TD course website that there is no use of market profile. Just trying to clear up some confusion. Thank you.

 

What Works in Trading & Why: Part 2
The Nature of the Beast

I have uploaded an annotated chart set for Gold. You will need to download that and unzip the file and pull the charts up to follow the narrative.

Trading Gold (@GC.P or GLD)

Chart 1 is a weekly chart of Gold futures. I trade both Gold futures and the Gold ETF (GLD) so the narrative here applies to both. Again, the analysis here is accurate, with the caveat that it is partial because I exclude additional proprietary indicators that facilitate the process.

At A & B (Chart 1) the same story: Momentum sellers had a variant perception versus value buyers. At A and B sellers were acting on noise as if it were information and therefore they were the noise traders -- and you can also see that hunting those noise traders has been a profitable thing to do.

As price retraced toward the bid line in mid-Feb 2006, at C, I was looking to get long because Gold had been bid up so well in early January that I thought the first reaction would be anxiously bid for -- but notice price did not retrace fully back into value at the bid line.

Now, look at Chart 2. I went long Gold 2/16 at A, and alerted the conference, but I was also cautious about this buy because price had not retraced low enough to attract a lot of value buyers so I was aware I may have been the noise trader -- too early and trading on noise as if it were information.

I took partial profits on the rally, but held a core long, but I never had a lot of conviction on the rally and had that subjective gut-feeling that price would revisit the low. The feeling was so palpable at B and I had a coincident sell on a different timeframe that I did a tactical short in Gold futures, at B, against my core strategic long. I took partial profits on the way down and also got stopped out of my core strategic long around break-even at C. I also prematurely covered my tactical short at C.

I had a pretty high conviction on the sell side for two reasons: 1) Price hadn't really pressed into the higher timeframe bid line (below value) at A in Feb and 2)
if Gold was going to minimize long side participation prior to the next rally (if one were to come), it was in accord with Hart's 7th Law ("That dramatic price movements tend to unfold from price structures that minimize profitable participation.”) probable that longs would get minimized by a thrust to new lows. So I put my tactical short back on at D, looking to cover that short on any thrust into the bid line at E on 3/20 (which I did), with the idea of reversing and reinstating a tactical long at E.

Referring back to the weekly Chart 1, at E, noise traders were prevalent and noise makes profits possible. Momentum sellers offering as Gold pressed below the Feb low were acting on the expectation of lower prices and the perceived probability of lower price was divergent from the actual probability creating a transient opportunity for value buyers to profit. The violation of the structural swing bottom (Feb low) had minimized profitable long participation by reducing core longs by bulls who had adjusted stops under that low so the conditions were ripe for long worth trying.

I did not reverse to long at E, like I should have, and I was a bit anxious about not being long as conditions in the higher timeframe were so ripe. I stalked the local price action and waited. Going into close on 3/23, I sent out an alert and got long. We're still long core positions in Gold as of 4/16/06.

In the Part 3, that follows (not sure exactly when) we'll apply the same process to another example but use current or out-of-sample data.

If you want my list of the exact actions we're now taking in stocks (and commodities) and my daily trading plan drop me an email at my blog:

Use the "Send me your BUY LIST" button, right-hand column at: http://www.street-noise.net/articles/

 

Hi Seth:

I thought Fisher's Logical Trader was worth reading and I think the principle that range expansion early in a trading day can offer useful trade location has merit.

The idea that range expansion begets more range expansion is arguable, or at least you have to try to position that idea in a context where the raw intrinsic value of any lower timeframe range expansion has enhanced risk-reward due to its trade location in some type of higher timeframe scenario.

His most useful idea to me, is the C-Reversal which in essence is the failure of a breakout to fully develop. In NoiseTrader Manifesto http://noise-trader-manifesto.com/ I speculate that what is not happening, often has more informational content for a trader than what is happening, so in the case of a Fisher C-Reversal the failure of an A-up or Down to develop and reverse into a C-Reversal has more information and I think Fisher implies something very similar in his book.

 

What works in Trading?

I'm still looking for that answer, but I think that the most successful speculators made their money betting against the ideas of the crowd. The successful speculator gets greedy when the herd is shocked by fear and gets scared when the herd is greedy.

Markets have an evil tendency to move in the direction which is going to hurt the most participants. Only a few make money while the herd gets slaughtered.

 

Hey Dan:

When I refer to Stedilmayer's methods, I am referencing anything the man ever wrote or said about markets.

Steidlmayer's work seems to me obscure, inaccessible and at times disjointed, he's no writer -- but I think that's our dim perception of what he has said.

He is IMO the most brilliant market mind of our time -- in fact so far ahead of our time that I doubt that he will be fully understood or appreciated in this century.

Hart's body or work lies in the Trend Dynamics course material and it contains little if anything of Steidlmayer.

(of course) I'm a big fan of Hart's Trend Dynamics -- I think the way he defines structure as an expression of market principles is innovative and helps to train the mind to see the underlying price action.

 

I opened an account here just to thank you for sharing your market knowledge.

You certainly have layed down ideas for the creative mind to run with.