|
|
发表于 2009-11-16 08:59 AM
|
显示全部楼层
Read this before you jumping into the bandwagon
A Balanced Review, October 30, 2003
By "hlcnprgrno"
Whoever reads the reviews for this book will note the polarity: people either think it's wonderful or it's horrid. I'll try to give a more balanced review here.
* This book is for the mathematically/analytically inclined. If you don't consider yourself in this category, I would not suggest this book. At the VERY least you should have a solid background in trigonometry.
* This book is exactly what is says: Digital-Signal Processing (DSP) applied to stock trading. I myself am an Electrical Engineer who's day-job consists of about 90% DSP-work / 10% other tasks. For this reason, I believe I can fairly say that the DSP concepts presented are done so in a fairly clear and fairly accurate manner (this varies from time-to-time).
* Most of the concepts presented in Ehlers' book would be what I consider "traditional" DSP. There is a problem here (that is almost completely overlooked in the book): "Traditional" DSP is based on a series of assumptions that must be at least approximately fulfilled. Recent academic research has tended to show that stock-market "signals" tend NOT to fulfill these assumptions very well. In some ways, this puts a significant dent (in my view) in the theory used as the basis for the indicators derived.
* The attention to detail at times is very obtrusive to a reader with a keen eye and understanding of what is being described. For example: there is (what I consider to be) a glaring error in the MAMA filter that no one has ever pointed out and that is not corrected on his website. How has this escaped notice for so long? [Side note: be careful about his claim that MAMA's avoid whipsaws - they DON'T at all in ranging markets!]
* Two mechanical systems are described in detail in the book: the SineTrend Automatic System and the ZeroLag Intraday System. I have EXTENSIVELY tested both systems, exactly as described in the book, on a very-wide range of stocks, commodities, and currencies. These simulations were carried out using a test suite I had developed for my own use to simulate trading conditions as realistically as possible. This includes conservative, but realistic allowances for slippage, commissions ($15 round-turn), and signal reaction.
Ehlers results for the SineTrend Automatic System grab your attention in the book (greater than $100,000). However, no initial investment is listed in the book and one should be careful to note that these results are over a 15 year period. Assuming a $10000 initial investment (though this isn't specified in the book), this is a 17% annual return. Not bad, not great.
The REAL problem here is that a more realistic simulation shows that one is hard-pressed to show ANY profit on ANY type of stock/commodity/currency. I've tried the system (unaltered) on dozens of instruments, with a couple rather moderate successes and the rest losing money terribly, even with reasonable stop losses and money management.
* Almost (if not all) of Ehler's graphs in his book are from the same time series: specifically the US96H (March 1996 Treasury Bond contracts). I HAVE been able to reproduce all of the graphs in his book using the indicators: so there is no falsification in them, as implied in other reviews. However, the time period under consideration shows a series with nearly ideal characteristics for Ehlers systems, while almost all other series I've tested his indicators and systems on do not exhibit the same performance.
* MOST ALL of the information presented in his book is available for FREE download on his website as Word DOCs. There is also a lot of other papers he has written available on the site.
* Ehlers is doing what I think needs to be done in the trading industry: trying new techniques and methods. He has a good grasp of the electrical-engineering concepts and how to apply them to trading, but the results presented in his book are NOT indicative of TYPICAL results of the performance of his systems. The ideas and the theory behind them are, generally, good, though he is a bit prone in parts to handwaving and glossing-over of some obvious pitfalls (obvious, at least, to another electrical engineer).
* For the mathematically-inclined who enjoy working on their own ideas and own systems, this book could give you a lot of insight into technical trading and provide some good jumping-off points to developing systems that are TRULY profitable. I have yet to see (or develop) one that I would be willing to risk real money on and I would caution readers from thinking that either of the two systems presented will provide the type of profits described in the book.
To all others (probably the vast majority of those interested in this book!) I would say be very careful before spending your money on this. There is no reliably turn-key system provided in this book, even though there are two that are purported to be such. |
|
狗仔卡
这里贴张图，是MESA的频谱分析和FFT（傅里叶）分析的对比，明显MESA的结果能得到最主要的波动周期，而FFT得到的结果糊做一团。这个mesasoftware还提供其他的指标，MESA是最贵的一个。另外还有很多
提升卡
置顶卡
沉默卡
喧嚣卡
变色卡
发表于 2009-11-16 02:26 AM
