you could not be more wrong (does not mean I fully agree witht he article);
Short term movements are mostly random, every beginner should know that and if not there is plenty quantitative analysis to support this. But I know you are a hard worker and want to always prove things for yourself: Alright, lets do it: Start at a 1-min bar frequency, establish a long if the current price is higher than 10 bars ago. Look at the performance 10 bars later. Do the same for shorts. Run various tests by varying the lookback and performance look forward parameter. Now step out and decrease the frequency until you get to a daily or monthly time frame. Report back what you have observed.
Next, market makers bread and butter does not come from order flow imbalances. Quite the opposite. Market makers perform best when they have a buyer for each seller and vice versa, resulting in no forced warehoused positions or additional hedging cost.
Please show me proof that market structure resembles fractals and even if it does (at this point I dont even know what you mean with it but please educate us) whats your argument and how does it point to support your claim that there are as strong short term as long term trends?
Moving into longer time frames does not introduce any additional variables in itself. If you mean that now fundamental factors also impact prices in the longer term then you are right but the increased number of variables in that sense have a beneficial impact on your predictive power not the opposite.
GSEs never had strong earnings power, everybody treated them as god-like entities because a paper tiger backed them. This has nothing whatsoever to do with survivorship bias, quite the opposite. Those companies with strong earnings power, innovative management teams are those companies who are still in the seat and have the cash to buy other companies. Its the Enrons, Tycos, and Lehmans that everyone is forgetting about and who either had horrible earnings power when time came to shut shop or they cooked the books. So, if you remove that bias then earnings power actually moves up couple ranks as price impacting factor in your PCA. Just a hint, but please do your own work to verify.
Before you start getting all excited about this article you may wanna think a little deeper in your analysis...just my 2 cents
Quote from FerdinandAlx:
The article makes some bad arguments.
1) According to the article the data shows that 1% of day traders are predictably profitable. They conclude from this that you'd be better of playing roulette. Yet 0% of roulette players are predictably profitable since it's a game of chance where the odds favour the house. This difference is meaningful as it shows that consistent profitability from day trading is possible, materially differs from a game of chance in a casino and is able to overcome commissions and other fees.
2) The author claims that short term movements in stocks are random. This claim is simply untrue. Short term movements are the result of order flow imbalances and there's an industry that profits from this called market makers. Furthermore trends can be shown to exist on shorter time frames as well as long time frames as market structure somewhat resembles the structure of fractals.
3) It's implied that short term movements can't be predicted but long term movements can. I'd say this goes against the nature of prediction which has it that predictions become less reliable as the number of variables increases. A longer time frame always increases the number of variables.
4) Investors should buy stocks of companies with strong and sustainable economic moats. While doing so is obviously not wrong the author does display a strong case of survivorship bias. Two of the companies the author names as lacking in this regard are Fannie Mae and Freddie Mac. Yet as recent as 2007 these two companies were considered to have just such a moat as both companies had a strong history of earnings growth and being GSE's they were essentially backed by the US government. In fact the article's disclaimer reads: "Rich owns shares of (...) Freddie Mac, and Fannie Mae (the latter two bought when he thought they had long-term competitive advantages)" Nice going there pal, I really hope that you'll be able to use the tax writedown to your "long term advantage".