Applying technical indicator knowledge outside of investing?

June 11, 2022 8:18 PM

A friend really likes using technical indicators for investing. Things like various moving averages overlaid on candle and bar charts for example. Since the markets are frustrating right now, this friend is looking for other areas where this knowledge may apply, and especially if they can learn about new types of indicators while learning about those new areas. They also mentioned throwing in some of those neat-looking & free math or statistics software packages out there, but are generally unfamiliar with them.

In case it's helpful, I know they like computer programming with really simple, retro-style languages like BASIC, and they use Linux a lot. Thanks for any tips.

posted by circular to Science & Nature (15 answers total)

Weather? Public health?
posted by NotLost at 8:24 PM on June 11, 2022

Response by poster: In case it helps: I'm hoping for recommendations like "ah, they use indicator X to perform task Y in field Z, and you could use software like A or B to do that.

Not sure if a big ask though. Thanks!
posted by circular at 8:26 PM on June 11, 2022

Real estate. Futures in things like grain. Professional art investing, though that's skewed by a lot of financial dirty tricks. Forex.

But TBH, if your friend thinks the current stock market is frustrating, they might want to check their models or faith in them. If your model only works when the market is going up (overall), it's not a good one. A good model can work as the market goes down, either by nabbing profit from market overreactions or through shorts.
posted by Candleman at 8:45 PM on June 11, 2022 [2 favorites]

Response by poster: Any non investing, non-money-making things? They are looking for a break in the investing.

I believe their methods work ok in markets so far from what I've picked up, but maybe they also really suck at it, hard to know for sure. Anyway...
posted by circular at 9:38 PM on June 11, 2022 [1 favorite]

Volunteering with a local progressive campaign as a data scientist? Maybe work with a animal conservation program helping analyze their numbers and trends?

Public health uses stuff like this but I'm not sure there's too much room or useful work that an amateur can do in it. There's a lot of uses for data science but only so many that will both expose their data outside the fold or accept results from someone they haven't carefully vetted.
posted by Candleman at 9:50 PM on June 11, 2022 [2 favorites]

Best answer: Binfo. DNA and protein sequencing is so cheap now that the pipeline is spewing out more data than all the computers and all the Effectives can hope to analyze in all of their lifetimes. But all that data is sitting in the public domain. I used to supervise undergrad research projects in a Tech Inst; students had never programmed anything and had limited computer skills but some made serious contributions to science. One set to looking at a database of Flu sequences and found better [than already published] predictors of mutations that would encourage viruses to jump the species barrier to humans from pigs or ducks. Another showed that standard members of the human gut microbiome had the pathways necessary to make neurotransmitters.
"ah, they use indicator X to perform task Y in field Z, and you could use software like A or B to do that." They discovered unannotated anti-microbial peptides AMPs in aardvarks, bats, chickadees, dolphins and on the regular using Blast. The software is written so that idiots innumerate lab biologists can use them.
posted by BobTheScientist at 1:55 AM on June 12, 2022 [2 favorites]

Lots of different markets... Stock market not well? Bond market? Commodities market? Derivatives market? Crypto market?
posted by kschang at 4:14 AM on June 12, 2022

Best answer: Charting-type investors like their models, but don't forget "all models are wrong (in the sense of incomplete), some are helpful."

We use flame graphs, percentiles and rolling averages in profiling and presenting dashboards of computer performance. We rarely estimate confidence intervals because 1/10,000 and 1/1,000,000 frequency events come pretty often with millions of events per day.

Install JupyterLab for workbook-like Julia and Python programming environment. There are lots of free data sets to try different analyses on.
posted by k3ninho at 4:49 AM on June 12, 2022 [2 favorites]

Best answer: Topics: economics beyond investing, the pandemic, weather, sports statistics especially baseball and football, politics especially election polling, traffic...

As for software, the starting point is Excel.
posted by SemiSalt at 5:12 AM on June 12, 2022

I don’t know enough about investing to give specific indicators, but a lot of finance types have gotten involved in sports statistics, to the point where a lot of baseball executives are former economics majors and/or Wall Street alumni. This can be just a fun hobby to understand a game, or you can use it for gambling to make money.
posted by kevinbelt at 6:25 AM on June 12, 2022 [1 favorite]

Best answer: Yeah, I’d think sports would be the go to answer for this, if they have any interest at all in the subject matter. Here’s the R task view for sports analytics packages - I imagine there is a similar set of packages for python:
https://cran.r-project.org/web/views/SportsAnalytics.html
posted by yarrow at 8:26 AM on June 12, 2022

They might find analyzing sentiment on social media interesting and the results can be applied to a lot of things.
posted by Candleman at 9:03 AM on June 12, 2022 [1 favorite]

Best answer: Your friend might find streamflow hydrology interesting. The USGS has a large network of calibrated stream gauges that provide near-realtime data on the flow of water through rivers throughout the United States. Your friend can almost certainly find a stream nearby, or pick a stream that has interesting dynamics.

There is a well developed theory of predicting stream flow from rainfall data. Some approaches rely little on knowledge of the physics of water, soil, evaporation, etc. These approaches are really applications of generic time-series forecasting methods also used in econometric modeling (and, I suppose, by "technical" traders). So your friend would likely find those approaches fairly familiar.

What makes this an appealing area for a hobbyist is that there is essentially no limit to how detailed a model one could build of water flux in a watershed, and there is a very objective measure of predictive success.
posted by phrontist at 9:48 AM on June 12, 2022 [2 favorites]

Best answer: Seconding Bioinformatics. If his interest lies in investing he surely is interested in quantitative investing. It's all about finding patterns in big data sets. He could practice this on this fantastic dataset, it's about a global project where scientists take swabs in subways around the world and sequence the DNA in those swabs. But they are not looking for human DNA in those samples, they want to analyze the microbiome. That means from the DNA sequencing information they can say which microbes were present and compare the compostion between those different locations.
I'm doing microbiome analysis in R and have no degree, so it's not that difficult to get into.
posted by SweetLiesOfBokonon at 11:33 AM on June 12, 2022 [1 favorite]

Response by poster: Thanks to everyone for contributing, this will be really helpful.
posted by circular at 6:44 PM on June 12, 2022