Post New

Does anyone remember the rules that are applied to a screen to make it a "Dozens" screen? I forgot.

Thanks, Larry

Does anyone remember the rules that are applied to a screen to make it a "Dozens" screen? I forgot.

Don't have the Dozens screen actually either. But running something similiar.

Count()                12
Schedule    12 Wk    DAYS()  Rebalance (Optional) or Replace
Basket1   01-Dec-25    -38   Next RebalanceIF HELD >= 12wk (84d)
Basket2   29-Dec-25    -66   Next RebalanceIF HELD >= 12wk (84d)
Basket3   03-Nov-25    -10   Next RebalanceIF HELD >= 12wk (84d)

GD_

Every month, you sell the stock you bought a year ago. Run your screen for the current month, and replace that stock with the highest-ranking stock you don't already own. If it's the same stock, of course you just hold on to it. It can get out of balance, and how you go about fixing that is outside the simple definition.

In a taxable account, you can fiddle with the timing, and sell losers at 364 days and winners at 366 days.

For starters, you can just buy the top 12 in the screen, and arbitrarily pick the first year's sales with something like lowest rank, or just pick losers (taxable account again). Or you could backtest it a few years, and start with what would have been the steady-state holdings from that.

In a taxable account, you can fiddle with the timing, and sell losers at 364 days and winners at 366 days.

Actually, that doesn't matter.

When you file your taxes, loses and gains are combined and you are left with just the net gain or loss and net sort-term or long-term.

Sort term loss of $1000 and a long term gain of $1100 = net $100 long term gain.
Sort term loss of $1100 and a long term gain of $1000 = net $100 short term loss.

Would be nice if you could take the loss at 22% tax and the gain at 15% tax, but it doesn't work like that.

The only way to do that would be to lump the losses in one year and lump the gains in a different year. That would be pretty hard to do, though.

Actually, that doesn't matter.

When you file your taxes, loses and gains are combined and you are left with just the net gain or loss and net sort-term or long-term.

Sort term loss of $1000 and a long term gain of $1100 = net $100 long term gain.
Sort term loss of $1100 and a long term gain of $1000 = net $100 short term loss.

It does matter, just not as much as most people would think. Throughout most of the year you cannot predict whether you'll end the year with a net gain or a net loss. So it's best to do what Jamie suggested. If you end up with a net gain, it's likely to be taxed at the long term rate. And if you end up with a net loss, it's likely to be deducted at the short term rate.

Elan

It does matter, just not as much as most people would think.

Well, yeah, it doesn't hurt, in the normal course of updating a screen, to sell a loser that you are going to sell anyway, to sell it just under the wire. and to delay selling a winner that you are going to sell anyway, to delay the sell for a few days to get it over the wire.

I think that as a practical matter, it only makes an insignificant difference.
I suspect that for most people they will be either keeping a winner much longer than a year or selling it much sooner than a year.

And if you end up with a net loss, it's likely to be deducted at the short term rate.
Not "likely". "Will".
Because $3000 of it will come off of ordinary income, which is the same rate as short-term rate.


I guess people like those doing heavily MI screens do monthly evaluations and sell stocks long before 365 days of holding. Both winners and losers. So it's rather a non issue.

The only opportunity you would have to play a few days before or after 365 days would be for an annual screen. And how many MI screens are annual?

Since short-term and long-term trades are netted out together, the only tax advantage there is for an excess of short-term losses is the $3000 you can deduct from your taxable income.

Basically, a short-term loss first takes away from short-term gain, then takes away from long-term gains. Leaving you with either a net long-term gain or short-term gain or up to $3000 deduction. The only advantage of a short-term loss is if you also have a large short-term gain for it to cancel out.

First you compute the total short-term trades, then you compute the total long-term trades. If both are net gains you have $XS taxed at ordinary income rate and $XL at long term rate.

If short-term is a loss, you subtract that from the long-term and if that is still positive you pay the total at long-term rate.

If LT is a loss you subtract that from ST and if that is still positive you pay the total at short-term rate.

If that sum is negative, you deduct it (max of $3000) from your ordinary income, regardless of if is net LT or ST. But it's effectively ST since it comes off of ordinary income. The excess loss gets carried forward to the next year. Where again the ST loss cancels out first ST gains then LT gains.

Where it *does* matter is if you can lump your ST losses and LT gains into different tax years. But even then the net effect is small.

It took me a few years to figure all this out. I focused on making sure my losses were short-term and my gains were long-term. Until I finally realized that it didn't matter in a hill of beans.

It's somewhat like the "tax loss harvesting" excitement. At first blush it sounds like a good deal, but when you look at the full situation over several years you see the effect is minimal.


--------------------

Throughout most of the year you cannot predict whether you'll end the year with a net gain or a net loss.

As I am sure you know, after you've been successfully investing for a long time most of your holdings are long-term and with large gain. So it doesn't matter if your losses are ST or LT, because they'll just reduce your net LT gain.

For the last many years most of my gains have been LT and most of the losses have been ST (and small). Because losers tend to drop out of the screen/portfolio fairly quickly.

Not in 2022, though. The ST loss was much larger than the LT gain. So all but $3000 of that loss was carried over to 2023.


Sorry, I didn't mean to go on for so long. Just delaying bedtime.

The only opportunity you would have to play a few days before or after 365 days would be for an annual screen. And how many MI screens are annual?

You're distracting from the subject of this thread, which is a Dozens screen, which does exactly that.

Sort term loss of $1000 and a long term gain of $1100 = net $100 long term gain.
Sort term loss of $1100 and a long term gain of $1000 = net $100 short term loss.

Would be nice if you could take the loss at 22% tax and the gain at 15% tax, but it doesn't work like that.

If the only thing you're doing for your investments is Dozens, no it's not going to be too useful. But if you have other short-term investments going, which are generally successful (as we all are, right?)...

Short term gain of $10000 on Monthly.
Short term loss of $1000 and long term gain of $1000 on Dozens = net $9000 short term gain, $1000 long term gain.
Long term loss of $1000 and long term gain of $1000 on Dozens = net $10000 short term gain, $0 long term gain.

So yes, it does work like that.

- Jamie

Short term gain of $10000 on Monthly.
Short term loss of $1000 and long term gain of $1000 on Dozens = net $9000 short term gain, $1000 long term gain.
Long term loss of $1000 and long term gain of $1000 on Dozens = net $10000 short term gain, $0 long term gain.


Sure.
If your trading screens are in sub-optimal locations.

Optimal location for short-term trading is in tax-advantages accounts. IRA and Roth IRA.

I just took a look at mine.
The "high-growth" (and possibly high volatility) screens are annual and in regular taxable accounts. That's things like VONG holdings and The "Magnificent Seven" and Top 10 of SPY.

The monthly screens are all in IRAs and ROTHs.

I did have a brain-fart a few years back and put a lot of B&H income-generating and high-yield funds & stocks into the IRAs & ROTHs, which I am slowly correcting.

Growth belongs in tax advantaged accounts, dividends belong in taxable accounts.


But, hmmm....
$9K ST at 22% and $1K LT at 15% = $2130 tax
$10K ST at 22% and $0 LT = $2200 tax
That's a $70 difference. BFD. On a $10,000 gain.
More than nothing, but insignificant in a total tax bill of $10,000 (avg for middle-class family in 2021).

But look at me, brushing off a $70 cost when I buy half-price used books instead of new books to save a few bucks.

-----------------

I guess a lot of it depends on how long you have been doing it and how large your portfolio is.
If you are just starting out, or don't have a 6+ figure portfolio, or don't have a good mix of taxable & tax-advantaged accounts then playing the ST loss game may have some effect.

$10,000 ST gain doesn't usually happen with small portfolio, though.

Post New

Mechanical Investing FAQ

---

Contact Shrewd'm
Contact the developer of these message boards.