Post New

Back in 2008 (post 208475) Jim wrote about lowering his cost basis. Is this a method you would still use?

https://yorickm.com/Message.php?pid=26523684
The secret is as explained in the post:
- pick anything that you are absolutely sure is below it`s intrinsic
value, and whose long-term-estimate intrinsic value is not going to
decline meaningfully (you can ignore one-off bad news, and in fact
it`s a good thing). The sort of business that, if it dropped by half,
you`d be perfectly happy buying twice as much. A good starting point
is any of the stocks owned by Berkshire Hathaway for a long time.
- Buy a chunk of it, say a given round number of dollars` worth.
- From time to time (anywhere from hourly to monthly), check to see
the value of your holding is above or below your original value.
Buy or sell shares as required to get back to the original value.
That`s it. You will vary your stock-to-cash ratio (or stock leverage
ratio) on each purchase or sale, in inverse proportion to how good
a deal the stock is. It works even better on multiple stocks, since
a certain degree of uncorrelation means you`re always moving money
from the lesser deal to the better deal. Yesterday I sold some BRKB
and moved some money into WFC, for example, since BRKB was up and WFC
was down since the last time I traded (last Wed, in this case).

For XLF, my buys have been between 22.82 and 26.61.
My sells to date have been between 23.22 and 27.27.
The sum of the cost of all my purchases less the sum of the proceeds
of all my sales gives my cost basis to date. This, divided by
the number of shares I currently own, is now $22.12.
Since the price is now 24.87, I`m up 12.42%. I entered this position March 17th.


I`ve been doing this on about 20 stocks since late last year.
The price improvement for that portfolio as a whole has been 5.74% so far.
The annualized rate (linearly, not compounded) is around 15%. This is
of course an unusually good rate because of the recent volatility.
This is calculated as a percentage of the target value of the holding,
which does not change, so your cost basis will eventually go negative.
This is the case for my Berkshire Hathaway holding, though it took
quite a few years.

DB2

Doesn't that necessarily mean that you have a sizeable chunk of cash initially? So that when/if the stock goes down you have money available to buy more shares.

And when/if the stock goes up you will be selling a winner? Why would you want to do that? Shouldn't you be keeping the winner?

The idea is that your PORTFOLIO increases in value, not that one particular piece of the portfolio goes up. The portfolio in this case is stock+cash. That's what you want to maximize.

The maxim is "buy on the way up, sell on the way down". This does the reverse. Sell as soon as it goes up. If it keeps going up and up -- which is what we want to happen -- and your portfolio asset allocation is going to be very cash heavy and very stock light. I don't much care if my cost basis has been reduced from $23.22 to $22.12 if my original 50/50 portfolio is now 90% cash and 10% stock.

Gosh, that was a while ago.

A short summary: It works, but it's not something I do any more. I still add to positions on dips pretty often, but nothing formulaic.

The main slight modification I made first is that rather than holding the market position value steady, make it so the market value of your position doesn't rise any faster than some estimate of the actual value of the stock.

e.g., Imagine you own and are happy with the prospects of $100k worth of XYZ (both price and your estimate of value). The estimated value of a share of XYZ then rises by 10%. At that point there is certainly no problem owning $110k worth of XYZ. It is not so great if you own $100k worth of XYZ and the price doubles in that same stretch. You now own $200k market value of a stock, a block whose value you estimate to be only $110k. That's a risk. The constant-position-size strategy cuts that risk a lot, as you'd sell $90k worth, in stages, for a good profit. It would still be $90k of an overvalued position, but that's better than $200k worth of an overvalued position : )

This addresses the concerns of not letting your winners run. They get to run, but no faster than their value rises. All price rises beyond value rises are ultimately transient anyway.

The strategy does definitely have the problem of having to have the cash to do buy more. I tend to hold a fair bit of cash, so that's not a problem, but looking at the overall portfolio return you would probably do better using other methods because of the cash drag. You could instead rely on broker margin, but that way lies madness. (and being broke).

In the end, I just found it too much work for the size of the benefit.

What I later found works much better as a way of reducing your breakeven over time on a set of core positions, and as a general portfolio strategy, is writing repeated cash-backed puts. I found that it gives a return of around the midpoint between [the return on the underlying stock in the same period] and [10-12%/year]. It also lends itself to a bit of extremely low risk leverage, since you never need $100k to back up $100k worth of notional puts on different stocks at different prices expiring at different times. I don't think I've ever needed more than 30% of the cash pile, and even if you did need more than 100% of your cash it would just be a broker margin loan for a day till you sell the stock and replace it with fresh cash-backed puts. The main drawback is that sometimes premiums are very low. You can see that in advance before opening positions, so you just don't do it at those times--invest in some other way. The original inspiration came from Jeffrey Cohen's book "Put Options", though I modified the strategy a lot. Mainly I don't add hedging formulaically as he recommends, only smaller amounts at times that the market seems unsustainably high and premiums are low.

Writing puts isn't free money, but it's darned close, at least if you are able to estimate the value of a few stocks. Getting back to the thought of the constant position size strategy you asked about: Say you're firmly of the opinion that XYZ is worth a solid $100 a share, where it happens to be trading today. If you already know that you'd definitely be buying more XYZ if the price dropped to $80, why just wait for that to happen (which might or might not happen), when there are people out there offering to pay you cash in advance to commit to doing something you'd already decided to do? Say the premium on offer is $5, so your net cost would be $75 a share. Being forced to buy something at net $75 a share that you already valued at $100 is hardly a loss situation, no matter what the market price is that particular day.

The very best candidates are firms that are definitely not going bust, but have sold off a long way because of some worry, but you're convinced the worry is fatal. You don't have to believe that the price will rebound any time soon. Such stocks can stay at low prices and high option premiums for years at a time, and you can make 15-20%/year from a flat stock price. They're often seen as duds, so you don't want to admit to your friends what you're long! I've even made good money on stocks that ultimately DID blow up in some way, but so slowly that I was making money faster than they were sliding. I made IRR 31%/year on USG before leverage on the portfolio, and 28%/year on Sears.

Jim

The very best candidates are firms that are definitely not going bust, but have sold off a long way because of some worry, but you're convinced the worry is fatal.

Oops.
Perhaps obviously, I meant to say
The very best candidates are firms that are definitely not going bust, but have sold off a long way because of some worry, but you're convinced the worry is NOT fatal.

Jim

Post New

Mechanical Investing FAQ

---

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