The term “indicator” always evokes a strong negative response from the retail crowd. If an indicator is just any transform of financial time series data, then indicators are practically what quantitative finance is built off of. 😭 #quant #quantfinance #trading #education

Institutions actually do use indicators and they do so in two main ways. They're not looking at SMA crossovers or trading off of the MACD or anything like that. Two of the real institutional use cases for indicators are these. One of them is feeding indicator-like transforms of price or other data into machine learning pipelines. That's where we're taking some time series data and computing these transforms like volatility ratio, realized volatility, volume Z score, etc. And we're feeding those indicators or those transforms into a machine learning model that's going to try to make predictions or give us signals. This is something that we'll never actually see. We're not going to see this on a chart. What we might see on a chart is model outputs. When an institutional strategy has a human in the loop, sometimes model outputs will be acted on by a real human being. For an example, that we're running a simple options pricing formula, which is this blue line here. That's the fair value of the option, whereas the real traded price is the white line for that option. And then the Z score of the difference between them is this orange line. So we have signals here like this would be a buy signal down here, another buy signal here, and these would be shown to a real trader on a chart to act on or monitor. Of course, this isn't an indicator in the traditional sense or anything that retail would be familiar with like the RSI or an exponential moving average, but it's the live output of a formula for a trader to make decisions based off of or monitor. So in a sense, that's a lot like an indicator.