This weekly roundup thread is intended for all culture war posts. 'Culture war' is vaguely defined, but it basically means controversial issues that fall along set tribal lines. Arguments over culture war issues generate a lot of heat and little light, and few deeply entrenched people ever change their minds. This thread is for voicing opinions and analyzing the state of the discussion while trying to optimize for light over heat.
Optimistically, we think that engaging with people you disagree with is worth your time, and so is being nice! Pessimistically, there are many dynamics that can lead discussions on Culture War topics to become unproductive. There's a human tendency to divide along tribal lines, praising your ingroup and vilifying your outgroup - and if you think you find it easy to criticize your ingroup, then it may be that your outgroup is not who you think it is. Extremists with opposing positions can feed off each other, highlighting each other's worst points to justify their own angry rhetoric, which becomes in turn a new example of bad behavior for the other side to highlight.
We would like to avoid these negative dynamics. Accordingly, we ask that you do not use this thread for waging the Culture War. Examples of waging the Culture War:
-
Shaming.
-
Attempting to 'build consensus' or enforce ideological conformity.
-
Making sweeping generalizations to vilify a group you dislike.
-
Recruiting for a cause.
-
Posting links that could be summarized as 'Boo outgroup!' Basically, if your content is 'Can you believe what Those People did this week?' then you should either refrain from posting, or do some very patient work to contextualize and/or steel-man the relevant viewpoint.
In general, you should argue to understand, not to win. This thread is not territory to be claimed by one group or another; indeed, the aim is to have many different viewpoints represented here. Thus, we also ask that you follow some guidelines:
-
Speak plainly. Avoid sarcasm and mockery. When disagreeing with someone, state your objections explicitly.
-
Be as precise and charitable as you can. Don't paraphrase unflatteringly.
-
Don't imply that someone said something they did not say, even if you think it follows from what they said.
-
Write like everyone is reading and you want them to be included in the discussion.
On an ad hoc basis, the mods will try to compile a list of the best posts/comments from the previous week, posted in Quality Contribution threads and archived at /r/TheThread. You may nominate a comment for this list by clicking on 'report' at the bottom of the post and typing 'Actually a quality contribution' as the report reason.
Jump in the discussion.
No email address required.
Notes -
Not much to say. I can't effort post but here's some rambling:
I don't know much math (and learned most in the last couple years) but architecting nice trade execution lets you do a lot of things; a good trade or correct insider knowledge is meaningless if you don't know how to isolate the opportunity from risks unrelated to what you want exposure to.
The systems themselves had a lot of inline assembly. Lisps all have some
disasmfunction giving a function's assembly, which you can improve and inline for easy high performance (or e.g. dynamically change). The architecture's were all OOP (i.e. moving hashmaps of data and assembly functions). In the same way a class can remove a level of if nesting, looking up particular fields in the object/map like "exit-inventory-below-optimal" or "exit-inventory-above-optimal" saves time, and those are all precalculated. They all used event sampling instead of time sampling. There were different models according to situation e.g. news can push the market to bimodal distributions around a new level, which governed particular data representation - laid out to encode decisions. Linear trend channels, vol compression breakout, support/resistance breakouts and trend change when linear trend channels break were insightful. I learned to write trading agents for each strategy (with an agent for each slight change e.g. for every .1% difference in stop loss) and all agents issuing internal orders, combined (e.g. some agents sell and others buy, canceling out) and then executed (Alan Dunne talks about "ensembles"). (I now only make a few trades a year with 2-5 year time horizons, so the agents' "votes" are weighted by success in the current and various other regimes, and they're working on various valuation schemes. Also log scale helps, because markets move return space.)Sampling is hard and important, since you need to choose data representations/current distributions/regimes etc. I like additive swarm systems. There was cool signal processing stuff for feedback control which I didn't understand, but which govern when to turn off (groups of) agents according to market stress and risk exposure. If you structure everything right, you'd have most computational power constantly rebuilding 100gb of hash maps and while the main loop does 2-3 look ups per agent on an event, everything on some group of correlated assets (like 5 gold mining companies in the same geography). (N.b. ensembles decorrelate things, different agents just with different stop losses have distinct return profiles even if only trading Brent.) Event sampling means if 20 things happen in an hour, but then 40 happen in 5 mins, and you're sampling every 10th thing... You'll have a lot going on during a little clock time on the spikes, hence precomputing things. Systemic indicators are driven by moving "windows" of data, whose updates are all recursively adjusted in the agent swarm. Remember, missing trades is fine but making bad ones is bad - so you'd have a more dedicated update loop for positions you're holding.
But everyone improved order execution, some HFT firms like Virtu shifted to providing order execution as a service. This is why IEX remains a small player.
Nowadays, off exchange trading/dark pools have similar volume to exchanges and while they're actually valuing assets, most can't see those transactions, which reduces overall price discovery. Far worse, passive inflows into indexes make up most exchange volume, which kills price discovery. You can do really nice things looking at the many thousands of stocks which have literally no analysts looking. (The investable world has really shrunk since the 70s, less quality markets (e.g. African and South American governments undertook awful policy so everyone left) and less publicly traded companies) but even the S&P 600 barely gets attention.)
Munger: "Investing is the only profession where inactivity is a competitive advantage."
More options
Context Copy link