Case study: Algorithmic trading with Go

I’m a practitioner and no, the costs to retail traders is included and analyzed.

The most vocal critics of HFT are very often previously practitioners who are upset when their strategies and models becomes obsolete are are outwitted by even more efficient operators.

Give any example of where you are going?

I will say RegNMS and NBBO regulations are actually preventing even further efficiencies. Dark pools and off exchange matching or internalization are complex topic that are easy to misunderstand. There are absolutely bad actors to be found in the system. This is true in any system. But on aggregate the system is continuously reducing costs and improving efficiency.

Intermediation costs are a friction on the real economy and capital markets and they will always exist, but on aggregate they are dropping for all users, institutional and retail.

In regards to party line, it’s absurd, you can simply take the aggregate income, not profits, of all the top HFT operators, their income is their counterparties costs, and when attributed per market, this number is continuously dropping on aggregate. If you look at an individual firm, you can see it’s income growing, however that will come from 2 dimensions, either expanding their operations to other markets, or taking share from a competitor. However if you sum all profits across all HFT operators on a single market complex, say US equities, on longer economic timescales, this number is continuously dropping. Obviously during periods of market volatility this number can increase, but the trend over years, will be always downward.

HFT is a absolutely brutally competitive industry.

I’d be interested in what you saw working on such systems seem “wrong” or “unfair”. My guess is you don’t understand that all is fair in war and that includes HFT, as long as it is legal.

I caught the algotrading bug recently and I've been wanting to backtest a couple of strategies involving ETFs. Right now I'm only looking at the open, close, daily high/low. The problem however is that you have to account for dividends and splits, which I'm not sure how to do (Admittedly, I haven't looked too deeply into it). Any advice on how to do that? Let's say I want to backtest some strategy involving UPRO; would I be fine just using yahoo finance historical prices as is, or would I have to manipulate it in some way? Also, what would be the ideal backtesting software/service to use?

It's a full time job bc your competitors are full time quant traders :)

No idea. I don’t know technically what is meant by this.

I am not disagreeing with you that the costs are dropping and that participants benefit from that. In fact I agree. Your projection that I find the field unfair is also unfounded. I love the field and find it immensely interesting. I just don’t take mythology surrounding it at face value because I often found it to be a) outdated b) full of mythology but no hard data.

What I am saying is that the thesis that the costs are dropping due to HFT style strategies has not been proven. Majority of the HFT tend to be market makers which tend to help with liquidity but not all are. I agree that liquidity helps offload or acquire large stakes.

As for NBBO and the effect a regulation has on markets you seem to be extremely US centric but if you go across the ocean and find out that Europe has no concept of NBBO at all, and that a retail person trading experience is equivalent for getting a different price on Amazon depending on which web browser they use you could imagine how that would make an average retail person feel.

Another example of the regulation is trying to move a large stake outside regular hours when NMS is suspended. Why do you think OTC block trades are pre-arranged at a fixed price? You can even look them up in relevant reporting facilities.

As for dark pools, ATS and internalizers there’s nothing difficult about them. I don’t think you’re doing anyone any favors by obfuscating an extremely simple concepts. Those market participants with their specific mechanics that are learnable. What’s complex is how to devise strategies and how to rely on the liquidity sources to get your desired fills and desired rates. You’re describing bread and butter of trading at an institutional level and the fact it seems difficult can maybe be attributed to the fact that it is opaque, doesn’t enjoy public communities, e.g. very little blog posts exist on the subject and the knowledge is sort of centralized to a specialized corners of the industry. But it is learnable without great difficulty if you have access to the resources.

What’s surprising to me is that a lot of people can talk about abstract concepts in computer science and then don’t connect facts that bridge into a separate discipline. As an example I could refer to consistency that people love to rave about but somehow forget the concept the moment capital market is introduced.

And yes, I’m familiar with the no trade theorem and relevant academic concepts but they are all models that often don’t translate to real life due to extremely limiting assumptions that don’t enjoy any connection to reality. I think it’s a problem with the community in this particular industry to take academic market models at face value. Often you will find that looking at a model carefully with assumptions that mirror real world, your performance gets better.