Voodoo Science-Statistical Analysis | Page 2 | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

[Book Discussion] Thomas R Demark - New Science of Technical Analysis - Chapter 1 30 replies

Best tools for statistical analysis of tick data and news events? 8 replies

Statistical Analysis of Trading History 9 replies

Statistical analysis fallacy 33 replies

The New Science of Technical Analysis book 3 replies

Attachments: Voodoo Science-Statistical Analysis
Exit Attachments

Voodoo Science-Statistical Analysis

  • #21
  • Edited 2:38pm Jun 7, 2016 2:11pm | Edited 2:38pm
Quoting Tardigrade
{quote} Median is supposed to be distribution-free??? Got any papers to back up your claim? Thanks
I was not referring to any paper. The median separates the data into two sets of equal size: the small values and the large values. If your distribution isn't symmetrical your median is neither the highest probability (the mode) nor it is close to the mean. I just see no practical use for it in this case. The median is good to get "a better mean" because it is robust to outliers.

Quoting Tardigrade
Nevertheless, a BB with 10 SD is worth a shot. https://en.wikipedia.org/wiki/Chebyshev's_inequality
A forecast with 10 standard deviations tolerence is like saying "With very high confidence EUR/USD price will go... somewhere". 100% acurate. 100% useless. Just my opinion.

Quoting Tardigrade
(pink noise) Why is that?
GPS devices give several readings per second. If you fix one on the ground and collect your latitude for 2 hours and you graph the samples you would expect a cloud of points around a mean value. This is not what you get. The picture below is the measurement of a GPS over 1 hour. The green line is the ground truth (they move their GPS in the experiment; their goal is to recover the green line). You can see noise around a curve which itself oscillates around the mean. Reminds you something?
Attached Image


EDIT: Picture from KALMAN-FILTERING WITH COLOURED MEASUREMENT NOISE FOR DEFORMATION ANALYSIS by Heiner Kuhlmann
Institute for Applications of Geodesy to Engineering, University of Stuttgart
No greed. No fear. Just maths.
  • #22
  • Jun 7, 2016 5:09pm Jun 7, 2016 5:09pm
Quoting Tardigrade
{quote}

1. Market returns are not normally distributed.

Quote
2. Statistics is a form of data science that collects, analyzes and interprets data. Falsely apply parametric statistics to market return analysis however, is "voodoo science".

Quote
3. The ridiculous example of measuring black swan using "sigmas" is exactly the counter-example of how faulted the original assumption of normal distribution could be.
OK, but there's more to it.
Quote
4. A "better" approach to market analysis is to use non-parametric statistics, median happens to be one of the most robust & simplest.
That's missing an important fact. The market is not a computer game based on some randomly generated time series. So statistical analysis, parametric or not, only goes so far.

Behind everything that happens in the markets are human beings, so before you get too deeply into statistics you'd better understand market micro-structure, i.e. how do the guys and few gals who move the market run their business. That'll lead you to study everything from Wyckoff Theory, to Market Profile, to how VWAP based client expectations have changed markets in recent years.

Having digested that, you may apply statistics to your insights, and then the parametric version is likely to be more than sufficient.
  • #23
  • Jun 7, 2016 10:13pm Jun 7, 2016 10:13pm
Quoting HeyYou
{quote} Assumption YES but not necessarily faulted.
Your rationale behind your claim?
No Highs? No Lows? Must be Bose!
  • #24
  • Jun 7, 2016 10:34pm Jun 7, 2016 10:34pm
Quoting mzvega
{quote} I agree................. This is why I prefer Auction Market Value Analytics. (Not a form of TA) It works using a frequency distribution; frequency distributions need not be symmetrical. If your data is distributed bounded by 3 price time events using only Market generated info/data. Market generated info/data is the markets transactional data collected and organized so that price and time are displayed. If you break the data down to its smallest measurable (quantifiable) measure the market consists entirely of price time events (units/tpo's)....
Thanks a bunch.

I'll study your original post "Auction Market Value Theory & Analytics", no rushy comments before I grasping the rationale behind it.
No Highs? No Lows? Must be Bose!
  • #25
  • Edited 1:49am Jun 8, 2016 1:06am | Edited 1:49am
I just see no practical use for it in this case. The median is good to get "a better mean" because it is robust to outliers.


A "better mean" leads to more accurate estimation of either location or scale; more accurate estimation of location provides more robust substitute for expected value. What to do with more accurate expected value? Think mean reversion strategies.

A forecast with 10 standard deviations tolerence is like saying "With very high confidence EUR/USD price will go... somewhere".

It can be backtested.

GPS devices give several readings per second.

Great example. Once again, if we take a few steps back and perform a logical check you can see that there is no "expected value" for Cauchy (Near Cauchy) distribution. Therefore an innovative method to find "expected value" is the first building block of a more powerful convergence system.
No Highs? No Lows? Must be Bose!
  • #26
  • Jun 8, 2016 1:24am Jun 8, 2016 1:24am
Quoting medici
{quote}That's missing an important fact. The market is not a computer game based on some randomly generated time series.
I agree. Before applying any form of market analysis one must first perform a logical check first; simplicity & complexity of modeling is not the issue here, the issue is whether a model could stand the test of time a.k.a. robust enough
No Highs? No Lows? Must be Bose!
  • #27
  • Jun 8, 2016 2:15am Jun 8, 2016 2:15am
Quoting Tardigrade
{quote} Your rationale behind your claim?

How do you explain that some traders are profitable after 200+ entries with a RR of 1:>1 and without martin gale ?

It's a proof that if we base our assumptions on solid past data we strengthen our probabilities. You may say "that's also an assumption!",
it is but still better than tossing a coin.

About the EURCHF... you call it "black swan", I call it DISASTER, trading pegged currencies is a no go and I'm sure many people knew it before the DISASTER.
  • #28
  • Jun 8, 2016 2:48am Jun 8, 2016 2:48am
Quoting HeyYou
{quote} How do you explain that some traders are profitable after 200+ entries with a RR of 1:>1 and without martin gale ? It's a proof that if we base our assumptions on solid past data we strengthen our probabilities. You may say "that's also an assumption!", it is but still better than tossing a coin. About the EURCHF... you call it "black swan", I call it DISASTER, trading pegged currencies is a no go and I'm sure many people knew it before the DISASTER.
Sometimes(most of the time actually) curve-fitted strategies perform reasonably well under one market cycle. Solid past data? Yes. Can the result be extrapolated? No.
No Highs? No Lows? Must be Bose!
  • #29
  • Jun 8, 2016 3:24am Jun 8, 2016 3:24am
Quoting Tardigrade
A "better mean" leads to more accurate estimation of either location or scale; more accurate estimation of location provides more robust substitute for expected value. What to do with more accurate expected value?
The expected value can be an irrealizable value. A dice roll has an expected value (and a median) of 3.5. Have you ever rolled a 3½? The expected value (and the median) of the return of a constant range bar is zero. It is impossible to have a doji in a CRB chart.

Quoting Tardigrade
Think mean reversion strategies.
Show me mean reversion in FX. I couldn't find any. One shall not confuse the recurrence with mean reversion. Brownian noise in 1 and 2 dimensions is recurrent. This means that whereever it is now it will come back after a finite time with probability 1. But it isn't mean reverting. Mean reversion means that some "force" makes the series come back to some value. Pulling it when too small. Pushing it when too large.

Quote
Great example. Once again, if we take a few steps back and perform a logical check you can see that there is no "expected value" for Cauchy (Near Cauchy) distribution. Therefore an innovative method to find "expected value" is the first building block of a more powerful convergence system.
Here the problem doesn't come from the moments of the distribution. The problem the picture shows comes from the fact that the noise is not independently distributed. The measurement error at time t gives information about the noise at time t+1. It is not Markov.
No greed. No fear. Just maths.
  • #30
  • Jun 8, 2016 9:35am Jun 8, 2016 9:35am
Quoting Tardigrade
{quote} Thanks a bunch. I'll study your original post "Auction Market Value Theory & Analytics", no rushy comments before I grasping the rationale behind it.
You may find some valuable insight here.....regarding different market theories. I think all traders need to know in which camp they belong, and why. I think one needs to know where their general assumptions about the market come from. i.e. many traders believe in EMH, which refutes TA, yet every trader at FF who claims to support EMH uses some form of TA.

The theory presented here is from empirical studies that provide logical, tested and well supported explanations for a number of observed facts
https://web.archive.org/web/20150526...ue_theory.html ......I think this link is best
https://web.archive.org/web/20151107...et_Theory.html
Markets are not efficient, rather they are effective - Jones
  • #31
  • Jun 8, 2016 11:39am Jun 8, 2016 11:39am
Quoting Tardigrade
{quote} Sometimes(most of the time actually) curve-fitted strategies perform reasonably well under one market cycle. Solid past data? Yes. Can the result be extrapolated? No.
What is the definition of a "Market Cycle"?
Wouldn't that require belief in an underlying periodicity in the market by imputing a regularity not supported by the data itself?
Again this goes basic to the issue of "Data Validity"

"A hope of many technical analysts is that markets have a cyclic component that can be used for predictive purposes. Belief in an underlying periodicity is implicit in moving average smoothing, since the moving average is expected to act as a filter. The same is true of oscillators, where two moving averages of different time frames are subtracted to find differences. The fact that oscillator differences often move smoothly and change fast in highly directional markets is taken as evidence that the resulting curves provide information that can be traded. An underlying but unspoken assumption of virtually all technical analysis is that market data is well conditioned (small errors in the data produce small errors in the result). In fact, as will be shown below, the fundamental assumption of essentially equidistant peaks in a set of market data is incorrect, and thus market data is poorly conditioned for most of technical analysis." - D.L. Jones
The study
http://www.forexfactory.com/showthread.php?t=572461
or
https://web.archive.org/web/20121229...tech_anal.html

The Market Unit
Auction markets are constantly seeking value. A perceived change in value will trigger a change in price. If value is stable (price is clustering) the market condition is 'balance'. A balanced market lists the center of value and the upper and lower limits of the balance. There is enough information there to make a trade setup. If the market is not in balance, market condition is either 'run' or trend or so choppy that there is no balance. Market condition goes from balance to run to balance to run and so on. The time from one run to the next is a 'Market Unit'. Posting the successive Market Units over a long period graphically displays the wide variation of one Market Unit to the next. This is also evidence of non-cycality in auction markets. - D.L. Jones
The study
https://web.archive.org/web/20121229024253/http://www.cisco-futures.com/marketunit.html
Markets are not efficient, rather they are effective - Jones
  • #32
  • Edited 10:39am Jun 9, 2016 8:24am | Edited 10:39am
Quoting mzvega
I think all traders need to know in which camp they belong, and why. I think one needs to know where their general assumptions about the market come from. i.e. many traders believe in EMH, which refutes TA, yet every trader at FF who claims to support EMH uses some form of TA.
If the statement: "A liquid market is efficient to all available information" is a form of EMH and a method of selecting trades using only price as an input uses a form of TA then the statement espousing a form of EMH is logically compatible with accepting that a form of TA can be used to trade profitably.

While I accept that liquid speculative markets are efficient to available information, I also believe that markets are complex systems where price is as much a cause as it is an effect. The act of price changing to reflect new information can itself be the impetus for changing perceptions of the underlying asset and thus the reason a market moves could simply be because it has done so in the past. This is evident in bull/bear markets that manifest in prolonged trends. With hindsight we may refer to these unpreditable events as manias or panics.

The form of TA that is effectively debunked by EMH is the naive form that asserts that TA can be used to predict the future. Markets are a system with as many different perceptions of value as there are participants and price at a particular point of time reflects the collective vote of all participants at that point of time saying nothing of the future. The speculator may accept that price has discounted all available information (hence, no risk-free profits) while accepting that past price could have an impact on future market events alluding to various behavioural theories.

And of course, I am obliged to add that many a price-based speculator would arrive at a profitable trading system by studying the works of Lefevre and Donchian and find optimal parameters for their system through testing... blah blah blah, boring stuff really.

Your questions are welcome.
  • #33
  • Edited 4:53pm Jun 9, 2016 4:38pm | Edited 4:53pm
Quoting Atokys
The form of TA that is effectively debunked by EMH is the naive form that asserts that TA can be used to predict the future. Markets are a system with as many different perceptions of value as there are participants and price at a particular point of time reflects the collective vote of all participants at that point of time saying nothing of the future. The speculator may accept that price has discounted all known information (hence, no risk-free profits) while accepting that past price could have an impact on future market events alluding to various behavioural theories.
Efficient Markets
According to Robert Hagin, The efficient market theory is the assertion that in a market with numerous investors who prefer high returns over low returns and low risk over high risk, information is of no value. An investor can attain no more or less than a fair return for the risks undertaken.

In the semi-strong form, the efficient market theory hypothesizes a market in which all publicly available information is efficiently impounded on the price of a stock, and hence a market in or item of publicly available information that can be used to predict future price changes.

An efficient market does not allow price to get away from value, since the theory holds that information is of no value since it is factored into the market, as every price must represent fair value. If price is always in equilibrium of value, then the theory presumes that prices are indistinguishable from one another in all but number. Thus, according to the efficient market hypothesis, all prices represent value, and the market is balanced at every moment: all the needs and concerns of the entire marketplace - participants in every known and unknown timeframe - are brought to the present by a price representing value for all. This seldom occurs and when it does, it is happenstance, and non-discernible except by the lack of activity.

Different participants have different timeframes for buying and/or selling; longer timeframe participants seek to buy low and sell high, and short timeframe participants must transact sooner and cannot be as discriminating. Thus, all markets including organized markets are inefficient from the standpoint that participants with differing time- frames view the most recent price with differing opinions and do not participate at the same time and at the same price. In other words, a given price is not viewed as an equal opportunity for all. The market cannot at any one moment balance all the needs and concerns of the entire marketplace participants in every known and unknown timeframe since the most recent price does not represent the same thing to all participants. In other words, no single price represents fair value for everyone.

Rather than being efficient, all organized markets are effective. They are effective from the standpoint that they produce and regulate trade, bringing into play all participants over time. In other words, the market facilitates trade from all participants, not with each individual price, but over a large sample size. The fact that participants with different timeframes exist precludes a price dominated by one time-frame from being fair value for all. The efficient market hypothesis overlooks the existence of participant timeframes.

The assumption of randomness and the efficient market hypothesis" assumes that the market - price movement over time -conveys no knowledge or information. Instead, it assumes that if a trader profits from a market, he has resorted to predicting, and his luck will sooner or later run out. However, one need not predict the future in a market of uncertainty. It is impossible to always accurately predict future events. But we can monitor the market, understand its present condition and how current price relates to the market's current assessed value, and from this, we can logically deduce probable behavior, and adjust our assessment from the new information the market provides as it unfolds. And by extracting bits of market-generated information, we can deduce and change. -J.Peter.Steidlmayer Markets and Market logic


Auction Market Value Theory
PR1: An example of a Non-permitted Practice, Efficient Markets:
It has been widely proposed that markets are efficient (any new information is immediately reflected in the price). This practice is invalidated by theory. Here's why: Markets are driven by feedback, therefore new information comes in at the feedback rate. Imagine that a given trader has new information and acts on it. That activity shows up on all trader's screens, say as an increase in price. Some buy on the news, some sell and some do nothing. It takes a different amount of time for each trader to decide. After some time (typically five to thirty minutes) the new information is assimilated and the market stops reacting to it. This shows market self organization at work, adjusting to the new situation. Feedback cannot take place instantly, so theory shows that efficient market concepts are invalid in auction markets. In plain english, even if a valid change in the value takes place instantly, the trader population cannot get the information, evaluate it and act on it instantly. - D.L. Jones
Markets are not efficient, rather they are effective - Jones
  • #34
  • Jun 9, 2016 4:56pm Jun 9, 2016 4:56pm
Quoting Atokys
If the statement: "A liquid market is efficient to all available information" is a form of EMH and a method of selecting trades using only price as an input uses a form of TA then the statement espousing a form of EMH is logically compatible with accepting that a form of TA can be used to trade profitably.
If price is always in equilibrium, reflecting all available information. as every price must represent fair value
Why would there be need for a TA method for selecting trades? If current price always reflects all available information, the current price is always the fairest value. If the market were truly stochastic and efficient, it should make no difference when the new stocks are bought and the old sold.

Why would one need a TA "method of entry" unless there is a underlying belief all prices are not valued equally?

In order to understand the market's inefficiency, consider the reality of how markets function. A normal preoccupation of the short time- frame participant is to focus on the dominant concern or news focus. The dominant concern causes the market to be balanced in the eyes of the shortest timeframe participants. It is the short timeframe participants who normally transact the vast quantity of total volume. But while the short timeframe participants transact up to 90% of the volume, their presence does not change the market. Thus, this dominant concern may or may not be a dominant concern for the longer timeframe participants, since it may or may not be of major significance to the supply and demand situation in the framework of the total marketplace. It is the focusing of the short timeframe participants on a relatively unimportant piece of news, a dominant concern that is not of major significance, that causes price excess. In other words, a minor piece of news a fact or perception that might be only 2% of the total timeframe decisions of the marketplace will motivate 80-90% of the day's or week's trade, since the short timeframe participants, who are the majority of the market, focus on it. Their preoccupation can create an opportunity when it is considered an excess for the longer timeframe participants. This leads one to wonder, how can all prices be equal, and how can all prices represent value when short timeframe overcompensation presents an opportunity for a more realistic valuation in a larger timeframe. -D.L Jones

Assuming that the theoretical efficient market exists in reality, it does render technical analysis ineffective. On the other hand, any evidence for the effectiveness of technical analysis is evidence against the validity of the efficient market hypothesis.........
Markets are not efficient, rather they are effective - Jones
  • #35
  • Jun 10, 2016 9:16am Jun 10, 2016 9:16am
Quoting mzvega
Efficient Markets
According to Robert Hagin, The efficient market theory is the assertion that in a market with numerous investors who prefer high returns over low returns and low risk over high risk, information is of no value. An investor can attain no more or less than a fair return for the risks undertaken. In the semi-strong form, the efficient market theory hypothesizes a market in which all publicly available information is efficiently impounded on the price of a stock, and hence a market in or item of publicly available information that...
You just copied and pasted that block of text wholesale.

Address the points directly, in your own words, or not at all. Lets try this again:

Statement A: Price in a liquid speculative market reflects all currently available information.
Implication of Statement A: Information discounted by the market cannot be used for the purpose of predicting future price with accuracy to obtain riskless profits. In context, a speculator cannot rely upon said information to obtain a risk-free trade by way of buying an asset that is "undervalued" by the market or selling an asset that is "overvalued" by the market.

Statement B: A speculator may expect to profit from a price-based trading strategy.

Assumption 1: Statement A on its own is a form of the Efficient Market Hypothesis.
Assumption 2: Selecting trades using only price as an input uses a form of Technical Analysis.

If Assumption 1 and Assumption 2 are true then a speculator who uses a form of TA is perfectly rational in subscribing to a form of EMH.

Notes:
- If Statement A on its own does not have sufficient qualities to be considered a form of the Efficient Market Hypothesis then we simply have a misunderstanding.
- If Assumption 2 is not true then we can cease the discussion as well.
- Adoption of Statement A does not commit one to supporting the full extent of views espoused by other EMH proponents.
Quoting mzvega
If price is always in equilibrium, reflecting all available information. as every price must represent fair value Why would there be need for a TA method for selecting trades? If current price always reflects all available information, the current price is always the fairest value.
My view is that price is a snapshot of information being digested into opinions. There is no equilibrium. The act of price moving to incorporate past information is already generating new information in the system that affects the course of the future in unpredictable but material ways. The value of an asset is the price at which it is exchanged. There is no implication that this value conforms to some standard of fairness, rationality or general agreeability. Im just assuming that this is what you are getting at from your language.

If a speculator studies the work of Lefevre and has grounds to believe that the course of past prices may affect future events in ways that are unexpected, unforeseeable and unknowable but potentially to his advantage then he is being perfectly rational and coherent in adopting both Statement A and B.

Speculators who operate in this manner do not forecast, they merely place bets. Most of these bets don't work out but a rare few exceed all expectations and win enough to make the overall venture profitable.
  • #36
  • Jun 10, 2016 12:07pm Jun 10, 2016 12:07pm
Since the OP mentioned using the median in place of the mean, an alternative measure instead of standard deviation is median absolute deviation or 'mad' in R. This effectively represents the average absolute deviation from the median (rather than the squared deviation from the mean as in the standard deviation) and produces statistics similar to what was shown in the picture in my prior post.

However if you subscribe to Taleb's brand of black swans it is important to remember that the deficiency of all these statistical measures is that the large unforeseen event, aka the black swan isn't in the data. So the statistics underweight the black swan's probability and potential importance - if it happens at all. In other words, you can't properly model data that isn't in your database. So doing modelling on data and extrapolating from this data to make predictions implicitly assumes that there will be no black swan type event in the future other than that which is already in the data.
  • #37
  • Jun 10, 2016 9:39pm Jun 10, 2016 9:39pm
Quoting PipMeUp
{quote} The expected value can be an irrealizable value. A dice roll has an expected value (and a median) of 3.5. Have you ever rolled a 3? The expected value (and the median) of the return of a constant range bar is zero. It is impossible to have a doji in a CRB chart. {quote} Show me mean reversion in FX. I couldn't find any. One shall not confuse the recurrence with mean reversion. Brownian noise in 1 and 2 dimensions is recurrent. This means that whereever it is now it will come back after a finite time with probability 1. But it isn't mean...
The expected value can be an irrealizable value. A dice roll has an expected value (and a median) of 3.5. Have you ever rolled a 3? The expected value (and the median) of the return of a constant range bar is zero.

Unfortunately, even if CRB produces near normal data, 3 sigma events cannot be treated as "normal", for traders cannot enter those trades in timely fashion(slippage, ping, etc).

Back to expected value, why would I need a realizable value though? Statistics summerizes data, can a single data point become "statistics"?


Show me mean reversion in FX. I couldn't find any.

The science behind statistics is irrelevent if we couldn't deduce any useful information from it. Have you ever consider asking the question as to why you can't find mean reversion in FX?

Spot FX market is quite different than futures, equities. It's OTC nature & without price limit mechanisms produced more 3 sigma events than any other markets you can find.


The problem the picture shows comes from the fact that the noise is not independently distributed.

If it were, it would effectively eliminates TA.
No Highs? No Lows? Must be Bose!
  • #38
  • Jun 10, 2016 9:44pm Jun 10, 2016 9:44pm
Quoting mzvega
{quote} What is the definition of a "Market Cycle"? Wouldn't that require belief in an underlying periodicity in the market by imputing a regularity not supported by the data itself? Again this goes basic to the issue of "Data Validity" "A hope of many technical analysts is that markets have a cyclic component that can be used for predictive purposes. Belief in an underlying periodicity is implicit in moving average smoothing, since the moving average is expected to act as a filter. The same is true of oscillators, where two moving averages of different...
Interesting Study. But I was referring to a different kind of market cycle.
Attached Image (click to enlarge)
Click to Enlarge Name: Cycles.png Size: 14 KB
No Highs? No Lows? Must be Bose!
  • #39
  • Edited 5:19pm Jun 11, 2016 5:05pm | Edited 5:19pm
Quoting Atokys
{quote} You just copied and pasted that block of text wholesale. Address the points directly, in your own words, or not at all. Lets try this again: Statement A: Price in a liquid speculative market reflects all currently available information. Implication of Statement A: Information discounted by the market cannot be used for the purpose of predicting future price with accuracy to obtain riskless profits. In context, a speculator cannot rely upon said information to obtain a risk-free trade by way of buying an asset that is "undervalued" by the...
It would be irrational for a speculator who uses a form of TA to subscribe to any form of EMH.

EMH has been proven to be invalid...

Rather than discuss the facts that support the invalidity of EMH, let's just "assume" that its valid and spend all day discussing "hypothetical assumptions" ignoring the invalidity of EMH despite the facts.

Price in a liquid speculative market reflects all currently available information, which means, is that the current price can never be "over or under valued". That would make using a price based strategy for selecting trade entry pointless. If price can never be "over or under valued" there is only ever one entry, that is current price. EMH is associated with random walk. I shouldn't matter when a stock is bought & sold. If the market were truly efficient, traders would disappear, as there would be no inefficiencies" to exploit.

The belief that one needs a price based method for selecting trades, flies in the very face of EMH. There is a reason EMH in all its forms refutes TA.

TA comes from a belief that some prices at times are more advantageous than others. These conclusions come from looking at past prices. Also a non-permitted practice in EMH. That assumption runs contrary to EMH.

Quoting Atokys
If Assumption 1 and Assumption 2 are true then a speculator who uses a form of TA is perfectly rational in subscribing to a form of EMH....
A speculator, who uses a form of TA for trade selection, is doing everything EMH says it's impossible to do. It runs CONTRARY. It would be irrational for a speculator who uses any form of TA to subscribe to any form of EMH. The act of using TA comes from a set of beliefs belonging to a different school of thought.

FF is littered with endless threads that go round and round discussing endless "hypothetical assumptions" not supported by anything but other "hypothetical assumptions". No one ever brings facts to the conversation. EMH is invalid. One of the several examples that support this, is that it overlooks the existence of participant tfs

Rather than discussing facts about its validity, people would prefer to go round & round continuing to "assume" its validity and discuss hypothetical scenario's around an assumption... despite the facts. It is a wasteful use of my time discussing hypothetical scenario's, based on an assumption...in light of actual facts.

The only thing that cuts through a BS assumption is a set of facts
Retail traders will always take the path of least resistance, it's too difficult to research a set of facts. So they talk in "assumptions rarely supported by anything more than another assumption.

So if I have to "wholesale" a set of facts into the conversation with a copy paste to untangle a faulty assumption I will. How much of my time & effort do you expect me to waste hand typing the same set of facts in "my own words"in discussions absent of facts where the knowledge base consisting of nothing more than a set "assumptions".

We could eliminate the waste of time going round & round discussing examples of an "assumption".. if we were directly addressing the facts

EMH is invalid and by adding an element (TA) the very thing EMH refutes, to the equation does not make EMH's invalidity somehow valid.

I dont have time to "pretend" that EMH doesn't overlooks the existence of participant TF's for the sake of a discussion of "hypothetical scenario's" based on an assumption not supported by the facts. It's called a "Hypothesis" for a reason.

It makes my point, every trader at FF who claims to support EMH uses some form of TA which runs completely contrary to EMH in all its forms.

This conversation is definitely not for me.
Same for using linear measures used to describe a non linear market, definitely not for me.
No more of my comments needed here

Good Luck & Happy Trading!
Markets are not efficient, rather they are effective - Jones
  • #40
  • Jun 11, 2016 5:42pm Jun 11, 2016 5:42pm
Quoting PipMeUp
Where is this thread supposed to go? It is known since the late 60's that the markets returns aren't normally distributed. Median assumes the distribution to be symmetrical i.e. no drift (aka trend/momentum). Mandelbrot has fit a Lvy distribution to the markets. This distribution has infinite second moment. How many "sigmas" (of Lvy) is this move? The answer is zero. Thus far I understand: There is no trend and the risk is infinite => markets are worse than casinos.
yes, market is cruel. its like slaughtering house
the gambler
the devils apprentice Return Today: na
0 traders viewing now