That's really not true. Most HFT firms are running strategies which are completely unrelated to anything a day trader would do, and most of the time they're helping you by being market makers rather than competing with you.

If what you said was true, no bank or hedge fund would run a trading desk. HFT captures just a slice of overall trading profits.

I'd argue that the reason why HFTs don't do day trading strategies, is because it's not profitable in the long run (ie, negative expected value).

Which backs GPs point.

Hedge funds engage in day trading, and there's many successful hedge funds. So the expected value can be positive if you're sufficiently skilled.

I always find these kinds of thought processes interesting "day trading always leads to losses". As with poker, anything which is predictable can lead to meta strategy which can be gamed: this is why GTO players can be outplayed by player-focused players, at least on occasion.

But occasion matters! Life isn't a nice continuous stream, it's lumpy. Circumstantial performance makes a difference - and that's why Greece could win the Euros despite probably not being a good football team.

Day trading is probably quite a hard discipline to follow, but that doesn't mean that everyone who does it is destined to fail.

Anecdotally 50% of marriages end in divorce but that apparently doesn't stop most people getting married...

Another way that I like to think about this is that if you think of a stock price as following brownian motion (e.g. follows a random, well defined process), then it follows that the stock price has a scaling-invariance property. In other words, the stock prices follows the same process regardless of the time window.. e.g. a day for day traders, milliseconds for HF, or years for long term value holders.

Now, brownian motion is just a model for a stock price so YMMV, but still an interesting idea. An investor can pick whatever time horizon interests them and that they're most suitable to take advantage of; e.g. HF traders take advantage of low latencies, technical analysis for day traders, and macro / micro economic analysis for value.

As the OP said, there are these three ways to look at this and at the end of the day, all are gambling with different time-horizons and this is possible because of this scale-invariance property.

I'm only correcting you because I'd want to be corrected.

Loose = Not firmly fixed into place.

Lose = Cease to retain.

Lose is the verb you want to use here.

Day trading is difficult and requires a lot of practice, knowledge, and patience, but saying you’ll lose every time isn’t always true. There are many many profitable retail day traders. HFT don’t really compete too much because they look to scalp differently.

A three to five percent success rate applied to a large population will produce "many many" winners. This is true.

You can apply the same logic to suggest no one should start a business.

That would be appropriate advice for a large slice of the population, so I don't see the problem there.

Business is a field where we know for sure that being better at it increases your chances to the point where your expected profit is positive.

Individual day trading is less clear. Sure, you can get better at it to increase your odds but it seems like that isn’t enough to make those skilled people in aggregate make money.

The analogy in a casino is that perfect play means you’ll do better but there’s still house advantage and so on average perfect players will still lose money.

Yes, if it's business is day trading. The success rates of other kinds of businesses vary.

And you have to be able to correlate their success rate with some measure of skill or else you’re basically saying that not everyone loses the lottery.

The claim to prove is that having some level of knowledge or skill makes the expected gains of day trading positive.

The people who make money tend to put in more money.

There are a number of different "day trading" strategies, some are easier and more reliable than others. Different strategies will work for different stocks, so it's really just a matter of getting to know the stock.

Big institutional traders are limited by risk, the fact that they are market makers for stocks in many cases, and the returns required need to be high, since their salaries are pretty ridiculous. Because of this, the fact that you're generally worse than them isn't a dealbreaker, just find a mid volume niche and learn it really well, then trade around general market volatility.

> I will lose against these guys every single time

It's true that you will not outperform HFTs consistently. But that's fine, you don't have to outperform them to make money. That's because the stock market is not just you and the HFTs, there are hundreds of thousands of other traders.

If there's enough liquidity, you and the HFT can make the exact same trade and have the exact same profit (percentage). The HFT is not necessarily against you. You might buy low and sell high to an HFT, and the HFT might sell even higher and make additional profit. Again, you both made money and weren't against each other.

I guess you mean High Frequency Trading, but I don't understand your reasoning. I'm an old school day trader as I execute my trades by hand.

HF traders DO NOT lose money. They see the trading world a few milliseconds into the future. The whole game is rigged in a way where they cannot lose. They only lose if someone screws up, human error, all that.

How do you explain all of the HFT firms that have gone out of business over the years? They're competing with each other, it's a very competitive space that doesn't even capture that much value compared to the people they replaced.

To win at HFT you have to have lower latency than your competitors. The NYSE rents server space in their building for this very purpose and makes alot of money doing so. If you don't have hardware there you will go out of business trying to compete with those who do. From a broad perspective it kind of looks like companies being allowed to skim freely from the market so long as the right people get their cut.

Getting early views of an order book and trades does not allow you to skim. No matter how close you are to the servers, you’re still seeing the data after the order has gone in and has either been matched or entered into the order book. There is no way for them to beat the order.

This is 100% not true. As mentioned below, HFT competes against itself, and there are measures to protect against latency manipulation.

I go from the account of Flash Boys where an HFT firm manager claimed they lose only once in four years - due to a human error.

You are wrong. HF traders front run to profit off individual trades by beating them to better prices. Day traders try to get an edge and exploit it to profit intraday by simply selling higher than they bought.

Front running is illegal.

Its payment for order flow which is earning off the spread while also keep it tight and liquidity in the market.

So... basically front running ?? That’s what it seems like you are describing. They make money when they manage to get ahead and provide the liquidity needed to fill the orders. What am I missing?

It is not front running, and can only be conflated with it if you don't know what front running actually is, don't know what PFOF actually is, or both. Firms that pay for order flow are paying to be the counterparty to transactions that have razor-thin arbitrage opportunities. They are not placing orders with advance knowledge of upcoming orders; they are reacting extremely quickly to fraction-of-a-cent spreads in the price of a security and pocketing the difference.

a market maker by law has to provide prices better than offered by the current market. frontrunning does not do that.

So you would rather your liquidity was provided by someone slower and more expensive?

Nothing. I can’t believe it’s legal. And yes, I know about the free trading it has supposedly provided. Maybe free trades are dangerous to to most retail traders?

And why would that be sustainably possible?

HF firms? Or pattern traders?

I’m not sure if my answer is good enough but the volatility is always there to exploit.