Three new functions: Rel%Chg , FutureRel%Chg , Future%Chg

Latest Release
,
1

Dear All,

We released three new functions. In the reference they are located under TECHNICAL->PERFORMANCE

Rel%Chg(bars [,series])
This function returns the total return relative to the benchmark (the default) or a specific series. This function should have been added a long time ago since it was very cumbersome to calculate, for example, the 1Y relative performance of a stock vs it’s industry.

Future%Chg(bars)
Returns the future total return

FutureRel%Chg(bars [,series])
Returns the future total return relative to another series (or current benchmark if not specified)

These last two functions should be very useful when we release the new API that makes generating data for AI/ML training simpler. The only use I can think of for these future functions using the website is to create fantastic simulated returns.

Let us know what you think

Thanks

2

Marco - Does this mean that FHist(“Rel%Chg(bars [,offset,series])”, -52) won’t work? Or is there some other reason for the future functions?

3

We don’t allow FHist as a data point in the API w/o a data license. We would have to look inside the quotes to see that it’s all technical data to allow it.

What’s wrong with using FutureRel%Chg(bars [,series]) ?

The main use case for FHist with negative offsets is for getting future fundamentals within your rules or ranks.

PS i made a mistake. there’s is no offset parameter for the new Rel%Chg

4

Cool, this will be very useful. I’d like to use this in Canada, are there any Canadian series I can use as a reference (ala getseries(“spy”) for the US)?

5

Never mind, I see that getseries(“xic:cn”) works great.

6

Marco - Three new functions: Rel%Chg , FutureRel%Chg , Future%Chg

Can you provide an example of how these functions might be used?

I don’t understand how we can know this information, considering how quickly things can change these days. I’m sure there must have been a need for it, but I’m at a loss for imagining what it might be…

I also don’t understand what you mean by this statement. Can you please enlighten us? Thank you.

7

The future functions will return NA if you run them today since we are not real time travelers. So they are useless for managing a live strategy. Their only real use is to generate data to train an AI/ML model. You give it past data (features) and future date (labels) to train the model to identify future winners

In a backtest they are just the ultimate look-ahead variable.

I should have excluded Rel%Chg. This one uses past data, and it’s very useful for backtests