Updates in his arrays are equivalently the same as changing his answers as time goes along. There are "many information" indicated by his array, and you MUST stick to measure the same thing everytime. If you say that his call on week3 is excellent in array1, and his call on week5 is great in array2, that is the same as cherry-picking the good results, while ignoring all the other bad results. So if in one array, you can derive 20 pieces of information, you should either measure the correctness of ALL information, or measure the same information systematically. You must stick to measuring the same thing over and over, whether the market is exciting or not. And for turning point calls, I would say that it would also have an above average success rate, if he simply call the turning point when the trend is about to exhaust itself indicated via the simplest technical analysis of RSI or MACD. Nevertheless, whatever you think he is good at, you need to measure that SYSTEMATICALLY.
And it's possible that you can grab out some above average results through this study, because after all, assuming that what he claimed about his past record on the Medallion fund is true, that indicated that he is probably a good trader (which is what I've been saying also). However, MANY people on Wallstreet had one or two-times wonders, but only one has been able to repeat his magic consistently (Buffett), although given billions of people on Earth, Buffett's record may not be even a statistical abnormality.
Try to do that, and let us know. What you measure must be quantifiable. It cannot be "I think he hits pretty close this time". You must define what's "pretty close". It must be in numbers. You can measure it by any delta of the prices, etc.
Armstrong appears to be "scientifically-minded", but he kept saying that the last peak was "confidence in government". If his computer can FIND the peak of confidence in government, then there MUST be something that was measured. But he just bla-bla-bla about it. Well, I think everyone who is watching the markets was sure that it should have been TLT (30-year US treasury), or any of the European bonds. The interest rates would have indirectly measured the confidence in government. I was so sure that Armstrong wanted to make that call of the century, calling the peak of TLT or whatever bonds. But it turned out that TLT didn't peak anywhere close to any of his dates. And so to make the story whole, he continues to say "confidence in government" without providing any quantifiable measurement. And of course, that's good enough for laymen. But again, if the computer can identify the peak, then there IS a measurement. Since he doesn't say anything on how he measures it, my guess is that he is making the whole thing up. Again, what he tried to do is to be the HERO here, calling the start of the bear market in the 30+ years bull market in bonds. Obviously, if I do the same thing, and call the peak every year starting from this year, I would have MUCH better chance of making a correct call, then say if I started doing that 10-years ago. Nothing goes up forever. The longer the bull market is, the higher the chance you can make a correct call.
And I'm well-versed in tech & science. Whatever he claimed on quantum computing is most likely fake, because as far as I know, you NEED to MAP a problem onto the quantum computing structure. It doesn't compute serially, and there is some structure that is required. You canNOT just "program" it. You need to map your mathematical problem exactly onto it, and then it can solve the problem for you.
For someone like Armstrong, assuming that he is the real deal, he shouldn't need to tell ANY lies. If he tells any lies, there is only one reason: to boost the confidence in him from the people whom he is trying to scam.
As I said, I'm guessing that he probably put out 98% truth, but just 2% untruth on his websites. To figure out that 2% untruth will not be easy.
Topic: Armstrong scams, TLT, etc. (Read 154 times)
Jump to: