I would like to ask a dummy question. Regarding processors and in particular the pulses per second of a microprocessor. The higher the pulses per second, the greater the power usage (energy/time) since you are converting electrical energy into kinetic energy. Computers at max power use 30 watts an oven uses 3000 watts of power. in simply terms Power=voltage*current. Power lost due to heat = resistance*current^2. So to deliver a given power efficiently you want a very high voltage (11kV) and small current . So it seems to me the clock speed of a microprocessor is really just about how hot can the computer get before melting at a given voltage. At 11 thousand volts not the domesticated 110DC i can design a microprocessor with a pulse rate of 188 zettahertz, keeping the current constant so it doesn't heat up. it is a popular misconception that clock speed is not important because of the amount you do in a cycle and the time for the new state to settle. this is false it is in fact of beyond critical importance if you are working sequentially and each cycle performs the most simple task possible. it allows you to sample the market state sequentially and continuously and always win (10 microseconds is for the birds, the problem of reacting to the start gun is a slow and long process and depends where the tick comes on the clock cycle). Parallel is not so much not the answer as an enormous step backwards as it is has all the same problems and far more fundamental ones, multithreading introduces so many problems it's hardly worth mentioning and is again mainly used to reduce heat at a given voltage. So I have hooked myself up to an 118gigavolt untransformed alternating current supply and am designing a pulser processor to clock at rates that would make a so-called supercomputer look like an abacus. 6ns is a standard clock cycle for a normal processor so even if everything else was zero you would have no chance just because of where you happen to fall. i would note 118 billion terahertz sequential sampling is a beta version, with additional supercooled superconductors we can easily approach pulse frequency on the plank scaled 5.39×10−44 s . It is conventional wisdom that fast has to be simple. This can now be thought of as false. Do you agree or not?
WHERE I PART WITH J.M. HURST I’m no engineer. I’m not even close. But unlike fellows like Hurst, who "claimed a 90% success rate trading on the basis of his theory," but about whom I can find no information suggesting he died a multibillionaire, or the brilliant mathematician Benoit Madelbrot, who "correctly used fractals to describe the volatility of financial markets, but could not offer analysts any tools that they could use to make money out of his findings," I fully intend to illustrate the utility of my approach to Forex trading, if it indeed exists, by operating with enough transparency to verify the kinds of earnings it is able to return in the real world. With this in mind, I’m not seeing Hurst’s fifth principle, the Principle of Synchronicity, pan out on my charts quite as described—that "waves in price movement are phased so as to cause simultaneous troughs whenever possible." What I’m seeing is both simultaneous troughs AND peaks, but perhaps that is because I am trading foreign currency pairs and Hurst was looking at equities. Also, it might be true that "there are an infinite number of cycles that…explain why it is impossible to forecast price movement with 100% accuracy," but that is not really the point, is it? The goal should be to verify when what has been predicted does unfold as forecast and to then trade accordingly. Moreover, my contention that numerical price prediction has "error growth with time" as an intrinsic predictability limitation applies to cycles as much as anything else, so that as one moves to the daily charts and beyond, cycles are of progressively less practical use, so that I have no interest in exploring them at such levels. And as far as the smaller time frames are concerned, once you move below one minute, the financial returns available to retail traders like me who are using cycles becomes so insignificant that, once again, the infinite number of waves that might exist are of absolutely no interest to me.
When looking into Hurst it might be helpful to consider that as an Aerospace Engineer he drew upon the science based analysis tools available to him in his research focused on wave theory specific to submarine and space communication and applied them to the non-science of stock market price movement. It is also important to consider that the data he had to work with was limited to daily, weekly, monthly OHLC and from a period when the markets were very different than today. What he found was a reasonable fit in that the science based wave principles seemed to account for the price action he was observing. I suggest the best way to take advantage of Hurst's work is to realize that it is only useful as an analogy in understanding price movement... it tends to explain price action more often than it does not. Some times it is accurate and sometimes it is not. It is also interesting to consider what his results would have been if he had today's sub-minute 24 hr / day tick data available for his research and was not limited by the processing power of a punch card data entry "super computer" with less processing power than today's smart phone.
You have this in your notes on your laptop, but just in case you forget about it at some future date and need to have your memory jogged to remind you it is there...
Monday / March 16, 2020 Sigma brought this entry from January 20th back to my attention today, which I am re-posting here so that I don't forget about it again when I am considering my entry about "Where I Part with J.M. Hurst." It just occurred to me that perhaps the reason Hurst only noted that "waves in price movement are phased so as to cause simultaneous troughs whenever possible" and does not mention them causing simultaneous peaks as well, as I am observing them doing, is because, like a lot of investors I know of, he was focused solely on buying stocks and not on shorting them. I also find it interesting that though Hurst held that "the cycles can be seen clearly if envelopes are constructed around the price movement," I am only seeing this tactic used by some traders, like Millard and Parisboy, but not by others, like Tillman and Eliades (though this might simply be because I have not purchased their books and the passages/images that deal with such strategies cannot be accessed online). Post #6:
I am fascinated with this stuff. More so I've began to focus my energy on the volatility of an underlying, more than the direction (although I pay much attention to direction too) The magnitude of moves +/-, i.e vol which is this immediate area which surrounds the spot price, seemed to be of more significance to my trading as time permitted. Especially when optionality was introduced to me some odd 7 years ago or so. Now instead of betting a binary directional bet with delta-one, I could strategize omnidirectionally via spreads across vol-strips. Pivot points are more useful to me at least, than something like Keltner channels. I was wondering @expiated if you use pivots in your technical analysis?
By clicking the "Like" tool, you caused the post to appear in my Alert box, which brought it to my attention.
The only thing I know about pivot points is this here... ...with Investopedia defining a pivot point as the following... a technical analysis indicator, or calculations, used to determine the overall trend of the market over different time frames. ... On the subsequent day, trading above the pivot point is thought to indicate ongoing bullish sentiment, while trading below the pivot point indicates bearish sentiment. However, from my perspective, if one has to set three different levels above and three different levels below price, it pretty much amounts to little more than guessing, and for that reason, I stopped trying to use them. Instead, I ended up using price range envelopes, but because I am using dynamic, adaptive price range envelopes generated based on historical data from the immediate past, I have found them to be both accurate and reliable IF used in conjunction with a valid baseline. Since they are not adaptive, I'm not a fan of Keltner channels, plus if I remember correctly, their usefulness in multiple time frames was severely limited.
Everyone is different and what works for one might not for others obviously. but you are very well-versed in technical analysis and I thought you’d use pivots in your own way. for example, pivot points are just the avg of the last days data, or whatever time frame chosen. This makes pivots a forward looking indicator, and that makes them much more useful then most rubbish out there. in fact aren’t envelopes, channels, bol bands just measures of magnitudes around spot price? @expiated can you elaborate on what you mean by adaptive envelopes? I’m assuming they are changing in real-time? But isn’t this the case with all technical indicators?