Really? So going from (100 x 20) to 5 x 1 or (150 x 10) to 15 x 1 is a problem for your mental calculations?
I always chart the contract value with that software. Maybe you didn't think about this, but the scripts running and deriving the overlay from the spread, have multipliers hard coded in the logic. Sooo, that's one reason.
the failure point is that humans are hopeful and that will often times taint the view of the facts. it's difficult to face failure after failure when it's so easy to just dream big and hope it works. m
There was a switching model where they identified three states of the market actually calm ,volatile ,and meandering within a range. Search for the paper on the quadratic rough heston process
It is actually superseded though by a process which is actually a continuum and not a discreet state switching model
I followed up and check one of so many papers. It was really super heavy on math. My question is, does this has any predictive value for trading or is simply just useful for calculating options prices?
actually steve777 is leading you in the right direction generally - don't be overwhelmed by the math get intimate with the usage of regime identification utilizing various methods of math. just identifying three states only will help any trader immensely.
yeah, I already know that, I use it a lot in my discretionary (paper) trading. Just looking for the easiest programmable way to express with code, with a smallest possible delay. In my eyeballing analysis I know in which mode I'm within 2-3 '5m' bars, but code typically needs like 5-10 bars. This code sluggishness actually looses money, IMHO. I'm just thinking loudly, maybe one wants to trade on longer time frame like '5m', but gauge state of the market on the faster time frame, like '1m'. The reason being that really high momentum moves have almost no consolidation within themselves, so '1m' and '5m' timeframes would give wildly different measurements.
There are a lot of noise in bars below 1H. Some traders probably would say even below 1D. If you try to tame 5 minute bars with any algorithm you will find that you end over-fitting your code to the dataset. If you are not after any sort of market making algorithm, 5 min will bring you more headaches than joy. And you are likely to chase those errors forever, noise is very hard to manage in trading.