I was just marveling at the supply chain for the dream liner for a school project. Your framing of the supply chain in relation to border tariffs makes me think Boeing and other globalize companies will rally hard against them. ]]>

Or do i check last week’s low and not just friday?

]]>This work would seem to lend itself to either a SVM or a Logistic Regression machine learning algo (ML). If you were testing this with ML which algo would you choose?

Thanks

Glen

]]>thanks Glen

]]>Glen

]]>You are right – the high/low range is a volatility estimator. Also, there is extensive econometric research on the time-series structure of volatility, for example the whole ARCH/GARCH family of methods.

But the proximity variable approach – which is what I am calling the methods used for these forecasts – is not closely related to these methods for modeling volatility.

So what I find is that the range – the difference between the high and the low in a period – is much more difficult to predict than, strangely, the midpoint of the range. I’m looking deeper into this apparent paradox, but it highlights the fact that these methods are not particularly good at forecasting the range, which is one metric for volatility.

]]>