I was able to do it consistently for over 3 years. I had a pretty successful strategy and could execute on that strategy using consumer grade trading APIs from Interactive Brokers. I had even extricated luck from most of the equation: ~9:1 profit/loss. It was an arbitrage strategy, so the luck variable became how often I had opportunities arise. I quit not because I couldn't make it work, but because I couldn't handle working alone for such a long time from a mental health perspective.
Here are my caveats:
1) low capacity strategies are the only ones that have a high enough profit/loss ratios but haven't been completely consumed by the hedge funds.
2) Because they are low capacity strategies, you can forget about compounding. Be happy with consistent cash flow.
3) Because they are low capacity, you can only feasibly scale across more products, not with bigger bets. This means that at some point, your connection bandwidth will become a bottleneck, even after you have graduated to professional data brokers. That then translates to a parallelization engineering problem. For me, it never made sense to deal with that complexity. It was better for me to have suboptimal earnings than to try to horizontally scale heterogeneously-active trading systems across servers in a data center.
Once you take into account the caveats, you can make a tech-industry level salary on your own with a lot of work. For some, the freedom may be worth it. For me it was, up until it wasn't.
It seems to me one of the bigger risk factors is attempting to maximize profit. If you're ok not trying to maximize profit it seems to me that you can mitigate quite a bit of risk in the short term trading.
"
You have a choice: $103k or your pride. To keep the $103k, you'll have to swallow your pride and admit it was all luck. If you choose your pride, you'll go to $0. The choice is yours. Best
"
Because no one believes a random person can find arbitrage opportunities that hedge funds or high frequency traders are not already exploiting.. In this case I think most people are wrong. I think it is possible to still find arbitrage opportunities in the open market.
Oh, you can find opportunities. They just don’t have the RoI and scalability that hedge funds are looking for. Sometimes bigger scraps reach the ocean floor and it’s a big day in bottom feeder’s life to stumble upon one.
Relying on constant insane returns isn't financial independence. Considering how many people are working to extract money from equities it's most likely luck.
Insane returns is one way to get to financial independence, because even if the edge exists, it's going to eventually disappear. You can boast about financial independence once you make enough to live the rest of your life from safe passive yield, after inflation, which in the current very low real rates environment requires a lot - maybe $3M as the bare minimum.
What you need to do is figure out the burn rate to sustain your family, figure out your safe withdrawal rate (e.g. 3.5%), then you can easily get amount of capital required.
A Mormon friend of mine, with 5 kids, spends less on food for his family than we do (family of 4) --he/his wife are just much better at spending on it. There are also factors like, it's possible to get a 3-4K sqft house for a very reasonable price & low taxes, vs. living in a tiny box in San Fransisco for 2x more.
Those historical returns are ridiculously optimistic. In fact stocks aren't great once you look globally. Imagine investing starting from 1900 Germany or Russia. In both cases you would be zeroed out. It's a form of survivorship bias to look at American stock markets only.
I think a 0% after-inflation return for a normal investor (ie. no specialized domain knowledge, no insider info) over the next several decades is already mildly optimistic.
This seems similar to Renaissance Technologies, especially regarding the regression part.
Funny how so many people are skeptical, as if they they want to be convinced that it's not possible. The fact Renaissance Technologies has been so successful for so long shows it can be done. Easy? no, but possible, yes.
Renaissance is a complete outlier. Most hedge funds - including firms that sell themselves as quants - just lose tons of money.
I've been in this industry for a while and folks making high returns solo either:
1) Got lucky for a very short amount of time, then their strategy stopped working. On avg those folks net lose money trying to make "the thing" work again.
2) Were running a super risky levareged strategy and went bust eventually
Even the Medallion Fund only averages 20-40% gains a year and they've stacked the deck with every possible advantage of resources and the most brilliant folks they can find. If the risk management on this is geared to the point where it's seeing 300% account volatility in a month, even in a positive direction, it's easier assign higher likelihood to
"this algo is levered far higher than it ought to be and while you've dodged landmines this month, you won't in the long run,"
vs.
"your 'minimum viable stat arb' and competent implementation outperforms the best-of-the-best outlier."
I think skepticism is warranted as most people at this point end up blowing up, but hey, I guess there's some slim possibility that it's the latter?
>Even the Medallion Fund only averages 20-40% gains a year and they've stacked the deck with every possible advantage of resources and the most brilliant folks they can find.
The Medallion Fund is also larger than a single quant's trading portfolio. This is important because a given trading strategy loses its effectiveness (in percentage terms) when scaled up. A strategy that yields 100% return on a 10k investment is unlikely to return 100% on a $10M investment, for instance.
RT is going to be a fantastic case study when their scheme unravels. The likelihood of the outcomes RT has enjoyed occurring without committing some kind of fraud or employing other illegal practices is basically (not exactly) zero. At this point it's more a question of whether their skill in hiding the shenanigans can continue to outpace discovery of them at a level of visibility that forces action by the authorities.
This is something that I'm passionate about. I'm about to launch https://feetr.io with the goal of trying to help people become financially independent through investing.
We measure momentum and use that to identify stocks with a good chance of massive price movement. It's currently free over at Twitter but will cost $8.99 per month.
Currently targeting people who know how to invest but long term goals are, of course, giving alerts to buying/selling opportunities and eventually investing on peoples behalf.
Historically the average stock has increased by 7.64% per day but given improvements to the algorithm we've managed 18.88% per stock this month. Though that does drop to 12.2% if you remove the 519% that COSM achieved.
All historical stocks are shown on homepage, with leaderboard at https://feetr.io/leaderboard (cuts the noise and gets to the exciting stocks)
I don't think sly would be required truthfully. The post is about becoming financially independent as a solo quant, I'm offering something similar. Currently for free. I'd like to offer the service to as many people as possible while it is currently free.
I know it's advertising and just about everyone reading the post should see it as such. That's not uncommon here.
Here are my caveats:
1) low capacity strategies are the only ones that have a high enough profit/loss ratios but haven't been completely consumed by the hedge funds.
2) Because they are low capacity strategies, you can forget about compounding. Be happy with consistent cash flow.
3) Because they are low capacity, you can only feasibly scale across more products, not with bigger bets. This means that at some point, your connection bandwidth will become a bottleneck, even after you have graduated to professional data brokers. That then translates to a parallelization engineering problem. For me, it never made sense to deal with that complexity. It was better for me to have suboptimal earnings than to try to horizontally scale heterogeneously-active trading systems across servers in a data center.
Once you take into account the caveats, you can make a tech-industry level salary on your own with a lot of work. For some, the freedom may be worth it. For me it was, up until it wasn't.