Post New

I have been working on and following a lot of MI technique for about 25 years, but haven't had any money in it for quite some time.
I just thought I'm mention that I'm restarting it Monday, since at least at the moment we're in a bull market.

It's very traditional MI, quite diversified, more or a "roll your own index" approach. First day will be 66 tickers.
Mostly boring stuff for most of it, similar to the philosophy of LargeCapCash, plus a few picks of potentially high returns just as leavening for the loaf.
For now the trading cycle will be every 2 months. Some things may have a longer minimum hold.

I'll post my returns from time to time, if people are interested. To make the tracking easy I put it in a separate account, and we'll pretend the starting value is $1 million.
The account will be paying 30% tax on almost all dividends received, immediately on receipt. Slightly different rate for a few things not US domiciled. No cap gains tax.

Jim

I have been working on and following a lot of MI technique for about 25 years, but haven't had any money in it for quite some time.
I just thought I'm mention that I'm restarting it Monday, since at least at the moment we're in a bull market.

Very heartening to hear that, given the shellacking that MI has taken the last 5-6 years.
I started back up a few months ago, somewhat tentatively to try to get the trading details down at IB, which is challenging.

Have to say many thanks to you. Just for the overall tone and quality of your advice over the years.
When things got rough with MI, twice I rotated into BRK-B which is now my largest holding.
My average entry price is something like $190 IIRC.
It has had an uncharacteristically strong runup lately and I have contemplated selling some to put into more aggressive screens.
That is surely a difficult thing to do. But seems like BRK-B will be cheaper at some point down the road.


Mark

Very heartening to hear that, given the shellacking that MI has taken the last 5-6 years.
I started back up a few months ago, somewhat tentatively to try to get the trading details down at IB, which is challenging.

Yes, IB can be a bit unfriendly at first, for sure. But I wouldn't go back for the world, now that I'm used to it. I'm trying the "basket" feature for the first time, but it's a bit scary.

As for MI screens, yes, there has been some "falling flat" effect in recent years. Maybe even more than the usual "screen was overtuned and didn't work after publication" factor. One thing I have found anecdotally is that final sorts were a big problem. A lot of screens that seem terrible in recent years are actually not bad if you look at a depth of 2 or 3 times what was originally intended. Also, dividend payers are out of their very bad stretch! For now, anyway.

But I never stopped developing and tracking screens, so the ones I'm running are ones the at seem that they might still have some value: they have held up for a while since invented. And I have very modest goals. If I can do a little better than the average stock over a decade, that's not so bad. By having modest goals you end up picking screens that aren't aggressively overtuned and they seem to hold up better.

Jim

By sheer coincidence, I just got some unexpected money from an early call of a CD (thanks, JPM, for yanking my high yield CD) so I am going to start the "one-step screen" mentioned in #1379, 3rd step variation.

The last one I tried (from the old TMF board) has worked out well.
The 90% closest to 52-week high, top 100 by ROE, top 25 by 5 Yr sales Growth. 25 HTD30 Trade monthly, no rebalacing.

Since Jan 2021, CAGR 13.54% vs. SPY 11.7% (all dividends reinvested)

Who know, maybe more MI screens will be coming back from the dead.

The 90% closest to 52-week high, top 100 by ROE, top 25 by 5 Yr sales Growth. 25 HTD30 Trade monthly, no rebalacing.

I don't know the specific post, and I would hate to claim someone else's idea, but that kinda smells like it has my fingerprints on it : )

A tip for a simple screen with a momentum step:
If you have the ability to test it, try it as an SOS of two or more variations of the screen, identical but for the momentum metric. e.g., one using price/high, one using price/low, one using 26 week, etc. The reason this seems to help (in smoothness, returns, or both) is that it does not react as strongly to any specific rhythm or frequency of direction changes in the market.

Jim

Yes, IB can be a bit unfriendly at first, for sure. But I wouldn't go back for the world, now that I'm used to it. I'm trying the "basket" feature for the first time, but it's a bit scary.


I have been using basket trader. That really seems the way to go for actively trading MI screens.
My orders are using their MidPrice algorithm and it does seem like the executions are good, but wonder whether VWAP may be better for less liquid stocks.
I do get a lot of warnings and notifications and it takes a bit to figure out whether to ignore them or not.
Also, I tend to have accounts that are 85% invested, so updating requires sell orders first with a delay for buy orders.
Otherwise, they will cancel an order or two due to not enough cash.
That seems to be happening even though I have margin accounts. Have not figured out why yet.

Ultimately I'd like to be able to enter basket orders in evening, or morning, and then go about my day without paying attention.
Not there yet.

My mostly manual process for updating right now is:
- export from IB the account holdings
- remove whatever holdings are not part of the MI blend (BRK-B and some other long terms holds for me)
- copy that portfolio holding list into a rebalance spreadsheet
- generate current MI picks and weights using either GTR1, RunRadis, or Radiscreen (in order of preference)
- import new holdings and weightings into rebalance spreadsheet
- decide if cash will be added/removed, and what rebalance threshold is (to reduce small trades)
- run macro that generates basket file
- log in to trader workstation and import basket file
- triple check that the orders make sense, $ values correct, etc.
- submit and cross fingers

It is much more tedious than the old FolioFN update and exchange method, and my spreadsheets are not elegant.
Hopefully I can automate this more using one of IBKR's API's, Python probably.


Mark

smells like it has my fingerprints on it

Yup.

[IBKR] is much more tedious than the old FolioFN update and exchange method

Look at M1 Finance. Their terminology is confusing, but once you understand the business about slices & pies it is easy.

That's what I use on that 25 stock screen. Run the screen, identify the stocks to sell and buy. Go into your "pie" and delete the ones to sell. Then in the pie add the ones to buy. The holdings allocations must equal 100%. You can click the "balance" button, or manually set your own percentages for the holdings.

Verify the sell & buy list, click OK. Go to the portfolio and verify the scheduled sells and buys.

The next trading window (next day morning) the trades will automatically happen in the portfolio(s) that are linked to that pie. Note that it will only sell the ones you deleted and buy the ones you added. If there is any money left over, it will distribute it by buying more of the current underweight stocks.

If you want to also rebalance your portfolio, go to the portfolio and click the "rebalance" box. Normally this will modify the scheduled sells & buys, so that it all happens at once.
Sometimes for some reason that doesn't happen, so you can just do a "rebalance" the next day.

I used to rebalance at the same time, but I don't anymore. I just leave the "keep" stocks alone. Let the winners run and let the losers (eventually) drop off. Equal balance is 4% allocation for each stock, currently mine run from 3.4% to 6.0%.

All signs are bullish. Jim is "back in" (another bullish indicator). Nevertheless, reading this on Saul's board:

... last year you told us that you weren’t interested in Nvidia because it was too big to 5X or 10x. But, here you are saying “ELF was a huge opportunity at $93,” but you sell out at $200? Just a double.... you’re looking for huge opportunities, but when they start working out, you abandon ship?

I have to wonder how much longer this bull can last with the infinite greed in "just double" so much smelling like 1999/2000.

I'm trying the "basket" feature for the first time, but it's a bit scary.


I've been using the IB basket feature for many years. Mostly when the markets seem very oversold.

I developed an Amibroker exploration that I run at night. This gives me a list of deeply 'oversold' stocks (not quite that simple really) that I copy into Excel and then export as a comma delimited file with all the headings that I want in the basket trader. The limit orders are at panic low levels (or that is the hope). If there are more possible trades than I want to invest in, then I put them in groups using the 'one cancels all' feature, so I can control the maximum dollar amount spent.

This is fairly automated and allows me to enter limit orders for all the stocks in the basket with no need to watch the market.

I use a short term bottom indicator in conjunction with the exploration.

Post New

Berkshire Hathaway FAQ

---

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