It's my education in statistics and the theory of information that makes me to believe that, not only the fact that I didn't find quants whose words I can believe. I'm a developer of a Forex simulation product, you can imagine how deep I went into the data to make it work. Technically speaking, all people who are trying to develop automated strategies use a set of the same small data files, several Mb in size each. We all know all possible algorithms that may work here, there is no secret formula. All these algorithms are based on indicators traders place on their charts, that's what makes them work, not some magic. You see big and complex charts filled with bars, lines, curves and marks which are built from really small static data files. GIGO (Garbage In - Garbage Out). So, from a mathematician/engineering point of view, there is a relatively simple and short noizy signal and a set of formulas you can apply to it to extract some meaningful pieces of information nobody else noticed before. If you think about it deeper, you will understand my skepticism. It may be fun to believe that you are The Chosen One, but I'm 38 years old to stop playing this game. I know that it was possible in the past. I found a guy who made $2,000,000+ out of $20,000 in 2001, using quantitative strategies alone based on trader groups' psychological profiles. I also know that all he did in the past can't be repeated at the moment because the competition is much higher in the quant universe. Where he made $20, you will make $0.002 nowadays. He was one of the first, now there are tens of thousands of software developers analyzing the same small data set again and again in various combinations. Even if you give me some proof that somebody actually made money using quantitative approach right now, will you bet on him in the future? It's a fun game, but I don't want to be a part of it. You may get more luck building strategies using some exotic data sets, like satellite imagery of parking lots, but how long will it take until everybody else notice this inefficiency? If we are talking about long-term perspective, human+AI will always win against any AI, no matter how complex it may be. If you are a strong believer in quantitative trading, you should be happy about my decision. Less competition is always better, isn't it?
@MaxPastukhov thank you for describing your opinion. What I read between the lines from it is that only one quant can use an anomaly ("edge") to make profit. And that this anomaly disappears when multiple quants start to use it. I am of the opinion that most retail quants do not have enough financial clout to occupy the space of an anomaly all by themselves. In other words: multiple quants could benefit, and make profit, by using the same anomaly. That makes the potential number of profitable quants much larger. There is an example of this on this forum: @globalarbtrader wrote a book on how to set up an automatic system to trade futures. And he shows in his journal that this system has been profitable over several years. In his journal you can read that several others started to copy his approach. And also they describe that they are profitable, using the same system.
Thank you for the link. Yes, you are right that there may be some inefficiences that may still be exploited at the market. They may even last longer than I may expect. But I would not bet my life on this thing. Imagine yourself back into 2001, when I had a chance to follow a guy that made 100x in less than a year. Will you decide to follow his system and ideas? Absolutely! The same guy traded differently in 2 years, before he died. I talked with those who personally knew him, who helped him to automate his trading. While he was eager to share, nobody of them achived the same results in the future. It's a different story, but you should notice that nothing lasts long on the market, only your personal skill and flexibility to market evolution. Even this guy, who was one of the first quants, traded manually. He didn't try to develop any kind of automated trading system, he always executed his trades depending on his gut feeling of the market. Someone once asked him why he closed one of his positions prematurely and the answer was: "There is a rule: I always close when I get tired of a position "
I do agree on this, that personal skill, flexibility and ability to adapt to changing markets (and changing opportunities) is important. However, I continue to differ in opinion with you on your original opinion that there is no future in quantitative trading. With apologies to @fan27 for hijacking his thread.
Good points...but dealing with more real world numbers. Trading Only $100,000 trading account 25% annual return. Year 1: $100,000 + $25,000 return - $60,000 (living expenses) = $65,000 Year 2: $65,000 + $16,250 return - $60,000 (living expenses) = $21,250 .......... Trading and Software sales Year 1: $100,000 + $25,000 return + $12,000 (software sales) - $60,000 (living expenses) = $77,000 Year 2: $77,000 + $19,250 return + $40,000 (software sales) - $60,000 (living expenses) = $76, 250 .... The above scenario is not my exact situation/projections as I have additional capital other than a trading account. As you can see, trading only is impossible for me because my costs are high because I have a family to support. If it was just me living very frugally, it could be possible.
I agree in that a quant is not going to be able to find big market inefficiencies that occur regularly and are highly profitable over time. However, being able to find numerous modest inefficiencies that occur semi regularly seems doable. For example, having a hundred strategies over numerous asset classes could provide a stable rate of return. The key is having tools for finding those strategies and determining when they are no longer "working". That is the goal of my product and its success depends on being able to do that.
What does being addicted to coding have to do with selling the product? Many times inferior product wins for meriad of reasons(some simply rediculous). Also, selling generic service (like web related dev) is not the same as specialized and highly subjective financial software. But if I had a product, I would probably try to see how far I could take it. Good luck.
Thank you! For me selling my own products is freedom, I did it for the last 15 years. I worked in a corporate world for the first 2 years of my career, dreaming about quitting every day. Since then, I earned not even a single dollar outside of my business. As for "highly subjective financial software", it all depends on the amount of time you are ready to invest into market research. 8-12 hours a day for 3-6 months will make you learn more about the market than your average customer, no matter how complex is the niche. It's how I came into the software business world as well. It was enough for me to read as much information as I was able to find on the topic for about 2 months to launch the very first profitable product. By doing it and taking short notes, then rereading them from time to time you can enter literally any niche. Yes, you will always be an outsider for top guys, but they will not buy anything from you, anyway. They usually have their in-house solutions or teams that are able to develop anything they want. It's good to get their feedback but you should always remember that your real customers will be different. The right strategy is to target beginners, not hardcore users or corporate ones. You just need to become a little smarter than your least educated customer. To tell you the truth, I was "an outsider" even in the previous niche after I decided not to do SEO in my business. I was a PPC guy but my customers were mostly from SEO/"black hat" world. But it wasn't an obstacle for becoming the leader in the niche for the long time. While this post isn't directly related to trading, I hope that fan27 will learn something from me As I didn't resist answering this time, I will add to his previous post that he is underestimating the amount of sales he can get in the very first year. Good launch usually takes 7-14 days, you just need to push the right buttons. In the past, I had launches from $20,000 to $100,000+ in a single day. I'm delaying it this time because I feel that there is something I need to do before the launch. It's not about the product, I will not even share it in public. PS: My biggest disappointment about software business from the very beginning was that it's not about how good you are as a developer but about how much time do you actually invest into market research and building relationships with others. At the very beginning of any business it's about 70% research/marketing, 20% communication and just 10% of coding. Yes, this number will gradually come to normal but only if you find somebody to do the same amount of research and communications for you It's really hard, by the way. Most people who call themselves "marketing experts" don't actually have a clue how to do it properly. They usually stick to a limited set of paid marketing channels, never investing their time into in-depth research. It's amazing how much you can learn about any company and the market in general if you just open and read every damn backlink to their website. You are absolutely right that sometimes crappy products win, you just don't know the inside reason of it. The hard truth is that the better you are as a developer, the harder it is for you to promote your products. I tried numerous times to convert all my friends into my "religion" but failed. All of them gave up when they found how much time they actually need to invest into market research and boring everyday marketing activities. 2 hours of market research can easily eat the whole day because it's really hard to focus and start doing it. When you see a mediocre product, it usually means that the author is either as bad in marketing as in product development, or the opposite. When somebody prefers to promote the product instead of adding new features and polishing, it's a good sign for the business. That's why I try to work in batches, focusing on one task at a time. If you try to develop the product in the morning, then promote it in the evening, or even each other day, it will not work. You are either focused on one clear goal, or fail. Software business isn't "fun" at the very beginning. It's fun only after you did a lot of really boring and hard work. Something like 2-3 months of prison, sometimes even 6-9 months if you are inexperienced and don't know where to start. This way you are replacing market research with trial and error. But then you can relax and tell everybody that it's fun, why don't they try it?
No, the day that i start selling my code must be the day that my code is not able t o make money anymore. It's all nice what you wrote and calculated, but reality is, people who know how to make and trade an edge, don't share. End of story.