TSPCalc Seasonal Builder

[onerepmax]

The "Builder" can only include full years. The "Calc" can include 2018, and will automatically include 2018 starting in Sep or Oct I believe.

[head]

Nope, all of 2018 is shown including today's gain/loss.
Thanks.

[LOCKER MAN]

Does the seasonal calendar change during the current year? In other words is it a live product meaning that my trading days later in the year can move +/- a couple days.

[fordest]

Does the seasonal calendar change during the current year? In other words is it a live product meaning that my trading days later in the year can move +/- a couple days.

No. Each strategy has set days that it trades funds. Those days stay the same no matter how the strategy is performing.

[LOCKER MAN]

Thanks. I was unsure about the trading days moving based on the market. As long as they are static that would be good as I took a screnshot o the caendar for my personal use.

[Rothwell]

Check this out: 45555

I created this strategy using only the G & S funds set for highest mean and optimized for the five most recent years (2013 - 2017). If you look before then you will see it is a mess but in the most recent five years it has averaged a mean of 38.64 and as of today (08-17-18), the year to date mean is 21.59.

[mjedlin66]

Check this out: 45555

I created this strategy using only the G & S funds set for highest mean and optimized for the five most recent years (2013 - 2017). If you look before then you will see it is a mess but in the most recent five years it has averaged a mean of 38.64 and as of today (08-17-18), the year to date mean is 21.59.

Based on that, check out Case 17:

https://tspcalc.com/seasonalblindtests.php

[Rothwell]

Well that is interesting, 45555 (the one I generated) and 48149 (from case 17) are identical.

[mjedlin66]

Well that is interesting, 45555 (the one I generated) and 48149 (from case 17) are identical.

Well that's not supposed to be possible. But I hear the guy who runs that site is a nincompoop.

[fordest]

I've noticed there are quite a few duplicates. I think at least some of them stem from that bug I reported to you, Matt. Here is one example. If you open my current Strategy: 31877 it opens just fine. But then click TradewithKup as a bench mark and hit submit. It changes to 46984. Its exactly the same as 31877. I am pretty sure it is the one that I created when I tried to manually fix that bug.

[mjedlin66]

Yes. There have been several changes in the back end that affected how the database identifies a strategy as unique. There will be another change with the next big site update.

It will all be cleaned up someday. Usually when I remove duplicates, I maintain an automatic forward. So if you look specifically for 31877, it would automatically forward you to 46984.

[Aitrus]

On that note, # 19323 and 45339 are identical. They might have duplicated because there's a glitch in how 19323 displays.

When the strategy is pulled up, the March allocations are kinda wonky. It shows a move to the S Fund on the 8th day, and the second move is to the C Fund on the 2nd day. When you look at the calendars below, it shows the C Fund move is actually on the last day, which is what the strategy should really be.

[Rothwell]

I noticed that the strategies I was coming up with that are built on the complete set of years are not performing well this year. When I created 45555, it was after a series of test building strategies with different set of year ranges.

It seems to me, that the strategies built only including the five most recent years (2013 - 2017) of data are actually performing well this year. For example, 45172, year range 2013 - 2017, all funds, optimized for mean has done 18.03 YTD.

Or 45553, with G + C + S, 2013 - 2017 optimized for mean has done 16.18 YTD.

There is something with that year range, maybe a shift in the market?

[fordest]

On that note, # 19323 and 45339 are identical. They might have duplicated because there's a glitch in how 19323 displays.

When the strategy is pulled up, the March allocations are kinda wonky. It shows a move to the S Fund on the 8th day, and the second move is to the C Fund on the 2nd day. When you look at the calendars below, it shows the C Fund move is actually on the last day, which is what the strategy should really be.

That's the bug I was talking about in my previous post that I PM'd Matt about. Any strategy that had a last day trade was converting it to a 2nd day trade. It messed up a lot of stratagies. Including mine. And #7980. Those got fixed so I thought they all had been. But maybe not. Hopefully his upcoming back-end update will fix the rest.

[Scarfinger]

Putting in Interfund transfer to "S" following daily strategy 19486

Waiting for site to update. Already made the change in my real account.