HYPTUSS Calculation change request

[vrdiver]

Whenever I update the prices, the Expected Dividend changes. That's as per design, as the dividend is a function of scraped price x scraped yield.

Would it be possible* to change the design so as to scrape the forecast DPS, and then use that to generate a calculated yield (which would vary as the price changed) and keep the Expected Dividend constant when updating prices?

This would avoid the over- and understatement of dividends when only updating the prices.

VRD


*Possible, maybe, high enough up the to-do list, I appreciate it may not be!

[kiloran]

Whenever I update the prices, the Expected Dividend changes. That's as per design, as the dividend is a function of scraped price x scraped yield.

Would it be possible* to change the design so as to scrape the forecast DPS, and then use that to generate a calculated yield (which would vary as the price changed) and keep the Expected Dividend constant when updating prices?

This would avoid the over- and understatement of dividends when only updating the prices.

VRD


*Possible, maybe, high enough up the to-do list, I appreciate it may not be!

An interesting idea, VRD. My initial thought is that forecasts have, by their nature, built-in inaccuracies, so we would probably be replacing one inaccurate number by another inaccurate number.

To get the ball rolling, let's take a simple portfolio of 1000 SSE shares. Looking at today's data from https://uk.webfg.com/equity/SSE, we get

Price: 1279.50
Forecast Yield 7.5%
Forecast DPS for Y/E 2018-03-31: 94.5p
Forecast DPS for Y/E 2019-03-31: 97.37p

Based on the current methodology:
Portfolio value: 1000 * 1279.5 / 100 = £12795
Forecast Dividend: £12795 * 7.5% = £959.63

Based on the new methodology:
Forecast Dividend = 1000 * 97.37p = £973.70

A difference of 1.5%, which I would think is insignificant compared to the inherent inaccuracies in the forecast.

The other issue is what data to scrape. Forecast Yield (if it exists) is always in the same place, in the Key Financials Income and Efficiency section. Forecast DPS is in the Forecasts section. But which row (which Year End)? For SSE we would (I guess) use the second row of data, for Year End 2019-03-31. But what about Diageo https://uk.webfg.com/equity/Diageo ? Do we use the first row (YE 2018-06-30) or the second row (YE 2019-06-30)?

And what about Investment Trusts such as Merchants https://uk.webfg.com/equity/Merchants_Trust . There is no forecast, though we could then use the last year displayed in the Key Fundamentals section.

So, lots of things to consider, and reprogramming effort, but but would the end result be any better than what we have? I'm not so sure.

Though I am sure that your post will generate some good and interesting discussion.

--kiloran

[vrdiver]

Thanks Kiloran for the response.

In my own portfolio, I use the "Expected Dividends" as a snapshot forecast for the year ahead (usually at end of December) and copy-and-paste the number so it's preserved and I'm able to compare it against dividends received a year later. it's one of the little things that gives me confidence in HYPTUSS that the numbers I'm looking at, at the portfolio level tend to be a pretty good predictor of what's going to happen!

I don't update the forecast dividends that frequently, partly because it's a little slower than the price update and partly because dividend changes tend to be much slower to filter through (anyone remember Carillion's forecast dividend recently? ). For that reason, I personally prefer the dividend figure to be stable, rather than fluctuating by whatever the share prices are doing.

If DPS isn't easily available, one option might be to scrape the current yield, as you do currently, but then multiply it by the current share-price and record that figure (a calculated DPS) which would then be used to update the Expected Dividends column, thus creating stability without having to figure out a new, consistent data element from the source.

In reality, I agree the day-to-day fluctuations are not such a big deal, but with the current system it's quite hard to see if dividends are increasing or decreasing, as the "previous" figure (i.e. the one you look at just before clicking "update" is not relevant.

All the best
VRD

[Itsallaguess]

If DPS isn't easily available, one option might be to scrape the current yield, as you do currently, but then multiply it by the current share-price and record that figure (a calculated DPS) which would then be used to update the Expected Dividends column, thus creating stability without having to figure out a new, consistent data element from the source.

Hi VRD,

Might another option be to construct another dividend column next to the YIELD-related calculations, and then carry out manual COPY/PASTE-VALUE processes from time to time, when you feel the need to take a 'snapshot' of the dividend amounts?

You could then change the 'Forecast Dividends' (total) cell to look at your snapshot-column rather than the yield-related amounts.

Would that get you where you need to be with the smallest amount of work/re-work?

Cheers,

Itsallaguess

[vrdiver]

Might another option be to construct another dividend column next to the YIELD-related calculations, and then carry out manual COPY/PASTE-VALUE processes from time to time, when you feel the need to take a 'snapshot' of the dividend amounts?

You could then change the 'Forecast Dividends' (total) cell to look at your snapshot-column rather than the yield-related amounts.

Would that get you where you need to be with the smallest amount of work/re-work?

Cheers,

Itsallaguess

Doh! It's the easy solutions that sneak right past me.....