OT bond funds

[mungofitch]

He owns crypto and doesn’t want to know about Buffett because he could never do what Buffett does because he doesn’t have a billion dollars.

Fair enough. It’s not an unreasonable stance.
There are many investments that are good picks for the portfolio at Berkshire Hathaway that aren’t necessarily good picks for an individual.
For example, some of the stock investments are like inflation-protected high coupon bonds with near infinite duration but not very much in the way of real growth.
For the portfolio of a biginsurance company, that makes a lot of sense. For me, not so much.
And all of their big positions are by necessity limited to being super large cap stocks. Again, not a problem for me. (unfortunately!)

However I think there is a lot to be learned by understanding why Mr Buffett does what he does.
My biggest insight is this:
Investing is just like the marshmallow test: Give up one now to get two later.
Once you realize that sooner or later every investment is about the future profits earned by the security in question, the only decisions in investing are:

• how many marshmallows you get later for each one you forego today;
• how sure you are that you’ll really get them; and
• how long you have to wait.
  The longer the wait, and/or the more uncertain the outcome, the more future marshmallows there better be on offer.

Any price change that is more or less than the change in value is going to be transient.
So everything else that people call investing boils down to trying to predict shortish term price changes.
That approach can certainly make money, and there’s nothing wrong with it.
But it’s a zero sum game against some very tough competition so it’s a much harder way to make a living.
Simple MI momentum doesn’t work very well in recent years.

Estimating plain old earnings, when you can manage it, works well even if valuations don’t change.
Random example: The P/E of Carmax (KMX) is the same as it was in early 2106.
But the return has 14.5%/year, simply because the earnings per share rose that fast.
(Obviously lots of firms have wildly fluctuating earnings so you have to be fancier, but that’s not true of KMX so plain old P/E works pretty well)
Nobody else had to be dumb and lose that money to you, and it didn’t have to get any more expensive.
So, in that sense it’s an easier problem…provided you can find some firms whose earnings you think you can predict at least passably well.
Off topic for this board:
Incidentally, at $97.68 at the moment Carmax is much (maybe 30%??) cheaper than usual at the moment, and one of my biggest positions.

Jim

[beginner3388]

Wow! Thanks, Jim. A lot there. Unfortunately I don’t know how to value things. Just Chapter 8 right now. I’d feel like I’m cramming. If I get the stuff I want within my circle of competence, then i might have time to study and use KMX as a lesson.

My friend’s husband, I feel for. Especially this very moment. he doesn’t have access to the minds on this board.

I learn a lot from you. Much I don’t understand and, for this individual, falls into the FOMO category, so I need to be mindful. Not in my circle of competence.

I hope to understand a bit about mechanical investing, which is why i read here. I need to read the sidebars but I haven’t yet. Meanwhile, it’s chapter 8, so I’m very watchful right now, sitting on some cash.

Bonds are hard for me, too. So I don’t touch.

Thanks to all the other posters, too.

[lintiness]

<Random example: The P/E of Carmax (KMX) is the same as it was in early ----2106----.>

The future is exactly why I read this board!

[DrBob2]

The P/E of Carmax (KMX) is the same as it was in early ----2106----.>
—
The future is exactly why I read this board!

Yeah, but did they mention what the earnings would be?

DB2

[mungofitch]

Wow! Thanks, Jim. A lot there. Unfortunately I don’t know how to value things. Just Chapter 8 right now.

Well, the big secret is that you don’t have to know how to value many things.
If they are really good firms, learning a dozen is enough. Even half a dozen would do.
(Though you might have to learn more to figure out which ones are really good firms)
My top three positions make up the great majority of my portfolio. I have a sense of how to value them.

Here is one to start you off.
Basically, I feed myself any my family by knowing how to value Berkshire Hathaway reasonably well.
I buy a lot when it’s really cheap, and (much less effective) lighten up when it gets richly valued.
And the value goes up around inflation plus 8%/year on average. More in good years.

The simplest yardstick of value is price to book. (best to use “highest to date book per share”).
This is the range of valuation multiples since the credit crunch, using price-to-peak-book:
0.928 absolute cheapest
1.164 10th percentile
1.274 25th percentile, today’s level
1.364 median
1.437 75th percentile
1.518 90th percentile
1.884 absolute most expensive

This isn’t a perfect metric, but you don’t need perfection.
(It has theoretical holes, but so far the holes have almost precisely cancelled each other out, giving essentially the same result as fancier methods)
Though I have spent years reading and studying and getting good at it, trust me that with this information above and little else, one can make a very good living.
You also need a serious ability to be patient, and some starting capital.

For the aggressive with a bit more experience, one can use that info to have a sense of when it’s
a good time to use deep-in-the-money long dated call options for a bit of uncallable leverage.
(never, ever, ever use broker margin loans)
I’ve averaged 25.2%/year profit on capital at risk in Berkshire for just under 21 years, even with the ending date at today’s fairly low price.
That’s after all the costs of trading and the cost of the leverage.
But I don’t pay income tax on realized gains, and “capital at risk” as a yardstick doesn’t include the portfolio drag from having cash on standby to buy on dips.

One other that I have tried to build a value yardstick for:
QQQE tracks the Nasdaq 100 equal weight index.
At $65, it seems to be trading at very near its historical average valuation multiple, which is not a bad deal.
I estimate “normal” value in the $59-64 range these days, which is close enough for rock and roll.
The advantages:
Being an index, it can’t go bust.
The earnings and therefore value have grown hugely faster than the S&P 500 for decades: about inflation + 8%/year, like Berkshire.
And, as it isn’t capitalization weighted, the company specific risk is vastly lower than the S&P 500 or the Nasdaq 100 “normal” index.
Each company is only 1% of the fund.

Jim

[mungofitch]

<Random example: The P/E of Carmax (KMX) is the same as it was in early ----2106----.>
…
The future is exactly why I read this board! ;

Yeah, I spotted that just after posting. Ooops.
I think most people figured out it meant 2016. I used figures from around the beginning of Feb.

Jim

[Inspired2learn]

Are there any resourcs which provides the price to book value in real time?

Or is there a way to calculate this ourselves in real time, based on the price which can keep fluctuating, and book value which I am presuming remains constant at least every quarter?

Thanks,
Charlie

[rnam]

Jim,

One other that I have tried to build a value yardstick for:
QQQE tracks the Nasdaq 100 equal weight index.

Elsewhere you have mentioned successfully investing in IHI - US Medical Devices ETF. If I recall correctly, you said investing in the triumvirate of BRK, QQQE and IHI, would have beaten the broader markets by a considerable margin.

What do you feel about the current valuation of IHI? At a forward PE of over 25, it still seems overvalued after a fall of 21% YTD. But then it always trades at high PE multiples.

I apologize if this is off topic for the MI board.

[mungofitch]

One other that I have tried to build a value yardstick for:
QQQE tracks the Nasdaq 100 equal weight index.
Elsewhere you have mentioned successfully investing in IHI - US Medical Devices ETF. If I recall correctly, you said investing in the triumvirate of BRK, QQQE and IHI, would have beaten the broader markets by a considerable margin.
What do you feel about the current valuation of IHI? At a forward PE of over 25, it still seems overvalued after a fall of 21% YTD. But then it always trades at high PE multiples.
I apologize if this is off topic for the MI board.

Actually I said nice things about investing in the stocks, not the ETF.
Most ETFs are cap weighted = bad.
Two reasons: too concentrated to do well (cap weight is bad), and in this case heavily concentrated in more diversified firms that aren’t as purely medical devices/supplies.
See post 278840

What I tested and recommended was a classic MI screen (so not off topic!)

Value Line set
Industry either MEDICINV “Medical Supplies Invasive” or MEDICNON “Medical Supplies Non Invasive”
Then pick a final sort.
Two suggestions were highest 5 year sales growth rate, or highest cash-to-market-cap ratio.

Elsewhere in the thread it was mentioned that having a Timeliness rating (any value) is useful, eliminating those in the middle of M&A etc.
And, optionally for more aggressive/higher returns, limit it to the ones not paying dividends.
(the dividend payers tend to bigger, stodgier, lower growth)

So, one specific recommendation was:
Remember that this is a VL screen, a slightly higher return 15-stock version could go like this
* those two industries, average 59 stocks
* ensure the stock has a timeliness rating (not too new, no huge M&A lately). This chops off 3-4 stocks on average.
* no dividend. this is the high flyer version! Down to 32 stocks.
* highest 15 by cash to market cap ratio

Since the final sort is not that critical, I lean towards “hold till drop” to cut down on trading.
So, for example 15 HTD 25.

As is usual, the screen underperformed the S&P badly in its first year after the post, by 24% (not annualized) in the stretch Jul-Nov last year.
Though it started outperforming a gain last September.
It made only 18% (not annualized) in the first 1.25 years after the post versus 25% for the S&P.
But I’ve been tracking screens using this industry for some years. 2011 or earlier.
This screen, 11 years to March (mostly in sample) would have returned 25.5%/year after friction versus 14.1% for SPY.
In backtest, anyway.

But, to your original question, what do I think about the current valuation levels?
I think it’s an industry that is a perennial outperformer, on average, but some useful amount.
(much less than you get in backtests, but a “real” amount)
As such, I wouldn’t put my fortune in it, but I can see a good case for having a section of one’s portfolio in it.
It will presumably go down when the market goes down, but it will tend to do better than average over time.
Unfortunately I don’t have a good rule-of-thumb valuation yardstick for this set, so I can’t really say whether it’s an auspicious good time or not.
One could speculate that a particularly good time is after a stretch of underperformance, like last December???
But if you want to have a look yourself at valuations, these are the current top 30 picks, in order:
SDC ADPT NVRO QDEL GKOS NVST TNDM ICUI NVCR IART
MASI DXCM AVNS NUVA SILK GMED OMCL HOLX HAE LIVN
ISRG BIO ALGN IRTC PODD ABMD CTLT PEN INSP VIVO

I don’t place a lot of weight on it, but FWIW, Value Line’s “Projected 3-5 year annual total return” averages 22.2%/yr for the top 10 of those.
One of the joys of the screen is its extremely low turnover.

Jim

[mungofitch]

Are there any resources which provides the price to book value in real time?
Or is there a way to calculate this ourselves in real time, based on the price which can keep
fluctuating, and book value which I am presuming remains constant at least every quarter?

Sure.
Price to book is on the Yahoo “statistics” page for both BRK-A or BRK-B.
https://finance.yahoo.com/quote/BRK-B/key-statistics?p=BRK-B…

But Yahoo is not always the most reliable source, so it’s good to know how to do it yourself. Grab a pencil!
Berkshire has two classes of stock; for financial purposes one “A” = 1500 “B”.
So, you have to calculate the number of shares outstanding.
Find the latest quarterly report here.
https://www.berkshirehathaway.com/reports.html
The number of shares outstanding at the end of the quarter is shown in Note 17 on page 19 of the most recent quarterly report.
(neither the note number nor page number changes much from period to period)
You can see that number outstanding at March 31 was 613960 “A” shares and 1285371832 “B” shares.
This equates to a total of 613960+1285371832/1500= 1470874 “A” share “economic equivalents”.
Then, go to the consolidated balance sheet, under liabilities near the end, page 3.
The line “Berkshire Hathaway shareholders’ equity” shows 508141 million.
(Basically, Berkshire has half a trillion in net assets, 7.5 times Apple and 3.8 times Amazon. Quite something)
So, book per share = book $508141000000 / shares 1470874 = $345469 per “A” share.
You only have to do that calculation once per quarter.

Then, just take the current price $439780 and divide by latest known book $345469 to get current P/B which is 1.2730.

If you do this for every quarter over the years, as I have, and also download the price history, you can calculate P/B for every day in the past.
I use ratio of current price to “peak to date” last published book.
And also see the progress of book value per share.
And, if you so choose, see the relationship between P/B and subsequent returns.
Sample results from that sort of exercise:
I inflation adjusted all stock prices, so the results are returns after inflation.
Dividing the daily P/B ratios since Jan 2008 into 20 ventiles, you get the following observation data for one year forward average returns:

0.930  to  1.150   18.6% in excess of  inflation
1.150  to  1.188   20.8%
1.188  to  1.227   16.5%
1.227  to  1.269   13.9%
1.269  to  1.301    8.4%
1.301  to  1.322    4.0%
1.322  to  1.337    2.1%
1.337  to  1.349    7.2%
1.349  to  1.360    9.5%
1.360  to  1.371   10.5%
1.371  to  1.382    9.0%
1.382  to  1.394    6.7%
1.394  to  1.408    4.3%
1.408  to  1.426    4.7%
1.426  to  1.442    3.5%
1.442  to  1.461    4.5%
1.461  to  1.486    1.9%
1.486  to  1.524    0.9%
1.524  to  1.583   -4.6%
1.583  to  1.930  -24.7%

I tend to do a curve fit through the data, then use the curve formula to estimate the most likely one year forward returns.
You can use other valuation models, and consider other amounts of history.
(Berkshire’s valuation averaged consistently higher before the credit crunch than it has since, which is why I used 2008 in the example above for conservatism).
For today’s numbers, depending on the model and time frame I use, I get likely one year forward return figures in the range inflation plus 12% to inflation plus 17%.
Those are just predicting AVERAGE returns starting from this valuation level. Averages can hide a lot of variation.
The prediction will be wrong, but the idea is that it’s a sensible central expectation: it’s a 50/50 shot whether it’s too high or too low.
Not all that many investments have a sensible central expectation of inflation + 15% these days, without requiring a big dose of optimistic enthusiasm.
It could be quite a bit worse and still be very pleasant.
I added a fair bit to my position on Friday.

Incidentally, how fast has book value per share grown?
Inflation + 10.1% since mid 1996.
Inflation + 10.5% since end 2008.
Inflation + 10.1% since end 2014.
Inflation + 10.1% since Q1 2018.
These figures are all a bit optimistic, since they have just had an unusually good couple of years creating an endpoint effect.
But inflation + 8% is a totally reasonable expectation. They might slow to inflation + 7% in future.
Book per share is not a perfect yardstick of value for Berkshire, but it’s in the “good enough” category.
Don’t try to use book per share as the yardstick of the value of other firms, other than investment funds or some insurance companies.

Jim

[branmin]

Just got a message from Ally. Savings account now is 0.90%.

Yes, I saw that and shifted two 11 month no penalty CD’s that were only giving me 0.55% into my savings acct. When I spoke to the agent over the phone who last night did the transfer I made it very clear when asking her would this be immediate and therefore lose no interest whatsoever and immediately obtain the higher interest.

First she said yes, it’s immediate, but when she read me the disclosure statement over the phone and asked me whether I agreed to this it included the part where it would take 1-3 days to process? So how do we know if it is immediate or not? Or indeed during the so called process, no interest is being obtained on the 11 month CD’s as those accounts are technically closed and in the process stage? I looked today at my account and it still shows the 2 CD’s and one savings acct!

Comments?

[esurfer2]

Thank you Jim for the post,

The simplest yardstick of value is price to book. (best to use “highest to date book per share”).
This is the range of valuation multiples since the credit crunch, using price-to-peak-book:
0.928 absolute cheapest
1.164 10th percentile
1.274 25th percentile, today’s level
1.364 median
1.437 75th percentile
1.518 90th percentile
1.884 absolute most expensive

For BRK stock one can find the Price/Book Value on a few sites.
1-What is the metric you use for QQQE? (price ratio?)
2-metric values for cheapest to the median to expensive?
3- Any website where one can look up this metric value

Thank you again for such valuable information.

[Inspired2learn]

Thanks a lot Jim! This is very helpful.

Charlie

[mungofitch]

1-What is the metric you use for QQQE? (price ratio?)
2-metric values for cheapest to the median to expensive?
3- Any website where one can look up this metric value

That’s a little more complicated.
Here’s what I did, which piles quite a few assumptions together.

Even though it is known as a home for high growth companies, many of which will have no earnings, ultimately it’s the earnings that matter.
You pay a lot for a high growth company because you think the future earnings will be high, not because of the growth rate itself.
So, though they might trade at very high average multiples (and quite reasonably so), it’s the trajectory of earnings that is the best yardstick of the progress of value.

So, the earnings yield is the best place to start.

Since it’s 100 companies equally weighted, the median earnings yield is a good approximation of what is going on.
It is much more numerically stable than the average or sum, because it isn’t affected by wild outliers.
You don’t really care about the outliers, since no firm is more than 1% of the capital allocation.
So, I used GTR1 to construct a data series for the median earnings yield of the Nasdaq 100 set over time.
You could stop here: by comparing the current figure to the historical range you’d immediately
get a first inkling of the valuation level.
You can look at the average-through-time of the median earnings yield, or the median-through-time of the median earnings yield.
Or just average those two, which I do!
In the era since 2005, which is after the tech bubble burst, this have averaged 3.925%. (P/E of 25.5)
The current figure is 13% below that, so the first rock and roll indication is that it’s 13% overvalued compared to “normal”.

But this isn’t great, since we noticed that the earnings have dips…better to use “on trend” earnings to smooth out the squiggles.
So we keep going.
Earnings are notoriously cyclical. We can’t do a cyclical adjustment without knowing what they were.
So far we have only the earnings yields, which is the ratio of varying earnings to varying prices.
So, we need a price series for the index.
I don’t know of a source for a the historical levels of the without-dividends Nasdaq 100 Equal Weight index.
So again I used GTR1 to build one.
Just tick the “Exclude ordinary cash dividends” box.
By putting this next to the earnings yield figures, and multiplying them, you get a data set of earnings on each date.
(technically, based on median earnings, but we earlier decided that was a good proxy for the whole set).
It is arbitrarily scaled, but that doesn’t matter.

Then I did an inflation adjustment on all those earnings figures and graphed them.
The main things that jump out at you are:

• The earnings dip a lot in recessions, more than you might guess, but rapidly spring back to their prior trend line.
  So, a cyclical adjustment really is necessary.
• Ignoring those dips, the trend line is an astoundingly steady progression over time. Good since 1997, but especially good since around Jan 2005.
• The growth rate has been very impressive indeed.

This was all done to get a way of estimating the cyclically adjusted earnings level of the Nas 100 equal weight index, in order to get a yardstick for its value.
So, I used Excel to run a trend line through the real earnings data set.
As noted above, other than dips during recessions, the long run trend has been remarkably steady, so we’re probably not making a wild leap here.

The implied earnings yield at the moment using the “on the trend line” earnings figure suggests an earnings yield of 3.75% because earnings are a bit below trend at the moment.
This suggests the Nasdaq 100 is 5% overvalued.
The price of QQQE is $65.05, but it “ought to be” (on the dodgy reasoning above) about $62.10.
That’s on the assumptions that:
(a) The average valuation level since 2005 is a not-bad estimate of future normal levels, and
(b) That the trend of earnings growth can be extrapolated without too much danger
Both assumptions will be off, but probably no so wrong that it makes the result useless.
Since this is not an exact science, $65 counts as “trading at fair value within rounding error”.

So, now you know how I came up with my estimate of fair value.
I have a few different models using slightly different assumptions (different mounts of history, more conservative trend lines, average rather than median etc).
I just redid those models, and get figures in the range $62-72.
With the price at $65, the indication is that it’s not parricularly overvalued, nor particularly undervalued.

Why is this an investment worth considering?

• It can’t go bust.
• The rate of growth of value is astounding. With the trend line above, earnings have risen at inflation + 8.2%/year. Plus you get about 0.5% dividend yield.
• It’s a simple one-step purchase.

Downsides:

• It has been overvalued a lot of time, so it has taken patience to get it at a reasonable entry price.
  It would be great to get it on sale, but that takes patience.
• It contains Facebook whose management I don’t like or trust : )
• It contains about four Chinese firms listed using legally unenforceable VIE entities with no secure ownership rights.
• The selection criteria don’t give certainty that the trend of high earnings growth will continue.
  They are merely the 100 biggest non-financial firms listed on the Nasdaq exchange, so they are growth firms only by tradition.
  The most conservative assumption would be t hat their performance would converge to that of the average firm.
  But that isn’t so bad, either, given the advantages above.

OK, as your reward for having read this far:
There is a shortcut.
As noted, at $65 it’s trading at around its fair value, give or take–around the historical usual multiple of trend earnings since 2005.
Fair value goes up over time with inflation. (I used CPI 292.3 in the analysis above).
Fair value goes up over time with trend value growth (about 8%/year)
So, you would probably be fine for a few years with this rule of thumb:
Take $65, increase it by the increase of CPI above 292.3, and then increase that figure by 8%/year since today, to get the new fair value.

Jim

[elann]

Yes, I saw that and shifted two 11 month no penalty CD’s that were only giving me 0.55% into my savings acct. When I spoke to the agent over the phone who last night did the transfer I made it very clear when asking her would this be immediate and therefore lose no interest whatsoever and immediately obtain the higher interest.

First she said yes, it’s immediate, but when she read me the disclosure statement over the phone and asked me whether I agreed to this it included the part where it would take 1-3 days to process?

Do you really care about losing three days of interest? If you invest $10,000 at 0.55% that gives you $55 a year. Three days of that income flow is about $0.82. BFD.

Elan

[mungofitch]

But if you want to have a look yourself at valuations, these are the current top 30 picks, in order:
SDC ADPT NVRO QDEL GKOS NVST TNDM ICUI NVCR IART
MASI DXCM AVNS NUVA SILK GMED OMCL HOLX HAE LIVN
ISRG BIO ALGN IRTC PODD ABMD CTLT PEN INSP VIVO

FWIW, meaningless spot check: those 30 returned an average of -1% in the four week hold.
SPY down -6.9%.
My portfolio down even more : )
That’s using the old MI conventional trading days: market close first market day of the week, close June 6 to close July 5.

The joy is the low turnover: only two changes to the top 30 after a month.

Jim

[Inspired2learn]

Hi Jim,

Are these based on highest 5 year sales growth rate, or highest cash-to-market-cap ratio?

I am assuming I am making a glaring error but it says TNMD has 5 year sales growth rate of -19%; and AVNS 5 year sales growth rate as -15% ?

Thanks,
Charlie

[mungofitch]

Are these based on highest 5 year sales growth rate, or highest cash-to-market-cap ratio?

Details in the post it’s a reply to, 283606
* those two industries, average 59 stocks
* ensure the stock has a timeliness rating (not too new, no huge M&A lately). This chops off 3-4 stocks on average.
* no dividend. this is the high flyer version! Down to 32 stocks.
* highest 15 by cash to market cap ratio

Ranking by sales growth is good too, but not the specific thing I was referring to.

I am assuming I am making a glaring error but it says TNMD has 5 year sales growth rate of -19%; and AVNS 5 year sales growth rate as -15% ?

Maybe you’re right. I haven’t checked.
So, some considerations–

• This particular slate was sorted by cash to market cap, not sales
• Remember that those were the picks from over a month ago. If new statements come out, the figures change.
• Value Line’s growth calculations are both a little quirky and a little error prone.
  As for quirky, I believe they smooth the annual returns then calculate the rate of growth of the smoothed line.

The important thing is that on average it has worked well.
In this particular case, it’s not that critical: buying the entire industry works fine, and changing the value of a sort parameter only changes the slate a little.
Especially since the final sort is still taking a large fraction of the stocks that are passed into it.
The top 350 stocks in the S&P 500 ranked by almost anything will not do much better than the whole 500 : )

Jim

[Inspired2learn]

Thanks a lot, Jim!
Charlie