Buffett buybacks: Could Berkshire tender stock?

[AdjustedEarnings]

Thank you to both the responses. You have given this particular issue more thought than I have. My question mainly arose because so many times we see something along the lines of... "the portfolio is down 2% since quarter end... so tax-effect and subtract that from the prior BV... add in earnings from operations, etc. etc."

 

We know that BV has been disowned by WEB... but, on further thought, I also have doubts about whether they use the MV of the securities. Not that they'd use IV either. But it definitely seems frequent adjustments would not be made based on weekly or even monthly moves in stocks. That would require moving the purchase threshold given to their broker (I think it's Citigroup) on a daily or weekly basis. I don't think this is happening (definitely not in a restricted period if they're operating under a 10b-5). So, to me, some of the very-fine tuning... ("Where has IV gone since quarter end") seems like work that would provide little additional information on which to base decisions. What do you guys think?

[alwaysinvert]

Thank you to both the responses. You have given this particular issue more thought than I have. My question mainly arose because so many times we see something along the lines of... "the portfolio is down 2% since quarter end... so tax-effect and subtract that from the prior BV... add in earnings from operations, etc. etc."

 

We know that BV has been disowned by WEB... but, on further thought, I also have doubts about whether they use the MV of the securities. Not that they'd use IV either. But it definitely seems frequent adjustments would not be made based on weekly or even monthly moves in stocks. That would require moving the purchase threshold given to their broker (I think it's Citigroup) on a daily or weekly basis. I don't think this is happening (definitely not in a restricted period if they're operating under a 10b-5). So, to me, some of the very-fine tuning... ("Where has IV gone since quarter end") seems like work that would provide little additional information on which to base decisions. What do you guys think?

 

Estimations of look-through owner earnings is the probable answer. My guess is that Buffett has some very simple, rough heuristic in mind, like owner earnings and if the whole company trades below say 20x of that he will be repurchasing stock.

 

I think that the multiple will be relatively high seeing as the additional cash flowing in from operations has basically no option value at all at this point. The bigger the cash pile grows, the closer the investment hurdle rate gets to that of the treasury yields. Now, he likely would never buy back stock at 40x just because treasuries were at an implied 50x multiple, but I think it is a useful way of conceptualizing the issue anyway.

[John Hjorth]

Estimations of look-through owner earnings is the probable answer. My guess is that Buffett has some very simple, rough heuristic in mind, like owner earnings and if the whole company trades below say 20x of that he will be repurchasing stock. ...

 

Again, we are mentally in synch here, alwaysinvert. I suppose that some staff member regularly provides him the numbers, or per specific & ad hoc request.

 

[Spekulatius]

Estimations of look-through owner earnings is the probable answer. My guess is that Buffett has some very simple, rough heuristic in mind, like owner earnings and if the whole company trades below say 20x of that he will be repurchasing stock. ...

 

Again, we are mentally in synch here, alwaysinvert. I suppose that some staff member regularly provides him the numbers, or per specific & ad hoc request.

 

I think we should use market value of the holdings, not intrinsic value. If the market value of Buffets shareholding’s goes down, very likely the general market goes down too and other stocks competing for dollars to be invested in BRK will get more attractive as well.

 

I wish I could mark up some illiquid stocks I own to fair value  :o

[Guest longinvestor]

My $208 filled. Hope my $202 does as well.

 

It did! I was not going to wait for 19x to happen.

[gfp]

I'm adding today as well.  Something like 500 Billion at today's price

[Jurgis]

I'm adding today as well.  Something like 500 Billion at today's price

 

Nice to hear that you have 500B to spare.  ;)

 

 

[bizaro86]

I'm adding today as well.  Something like 500 Billion at today's price

 

Nice to hear that you have 500B to spare.  ;)

 

Probably want to use limit orders on that one.

[Dynamic]

To the question of how Charlie and Warren value the investments, I don't think it matters so long as they are suitably conservative and are not in excess of market value.

 

Using Market Value itself or applying a discount to Market Value to apply a windage factor to normalize to a more typical market valuation is one approach.

 

If they valued the companies independently they would either be calculating IV with a high enough discount rate or capitalization rate to be conservative or they'd calculate a more fully-valued IV then apply a suitable discount to be conservative. (And in a similar way they would be valuing subsidiaries with a similarly conservative assessment of what they're worth given the current normalized earning power)

 

Whichever approach they take they'd arrive at a similar figure, being conservative, I'm sure.

 

At the end they may choose how to account for cash held and how to value float and normalized underwriting profits.

 

A few commentators have noticed that over the years the cash balance is usually pretty close to the float liability and only dips significantly below float during bear markets or similar opportunities to make large acquisitions, and that remains roughly the case today. To me this offsets the feeling that there may be a cash drag if the cash doesn't get invested fast enough, but instead limits the multiplying effect of the float leverage except when cash is being used up to buy cheap assets.

 

Float is a liability, but the funds are so likely to endure (3% per year decline rate at most, but more likely to gradually increase over time) and to remain cost-free (thanks to profitable underwriting) that one could effectively count as much as 70% of float as an effectively enduring asset whose economic earning power is worth paying for. Perhaps, being more conservative, we'd value it at less or apply a margin of safety at the end of calculating the full IV.

 

To be frank, I'd expect them to make a relatively simple calculation with amply conservative assumptions built in rather than going to great lengths to adjust market values of securities very much. They're still intending to pay significantly less than IV, but perhaps a 5-10% discount to a conservatively calculated IV amounts to the same as a 30-50% discount to an IV that represents the line between fully valued and overvalued.

[SwedishValue]

Buying today gives you a quite decent discount to the price that Buffett started the repurchases. With increased volume during this October turmoil, Buffett has been presented with the option of buying back significantly more stock compared with the previous months. Not that it protected the downside that much (I didn't expect Berkshire to drop more than the market during the so far worst parts of the turmoil).

 

In this type of market environment, will Mr. Market approve or disapprove of a communicated and big Buffett buyback? A month ago, I was very confident that it would be the headline. Today? Not as sure.

[valueinvesting101]

Positively surprised by Berkshire fall too. It is good since buyback can happen at an attractive price.

 

It is all speculation but it could be triggered by Index funds or charities selling stock to meet yearly spending requirements too.

[SwedishValue]

When things goes sour, everything turns out to be correlated. This should be a nice opportunity for long-term shareholders.

[Swedish_Compounder]

I realize that some members might be tired of the question how to value float and float increases that comes up from time to time. I just wanted to share WEB:s Words on the topic, since it shows that his opinion is that float is worth more than equity (as long as it costs less than zero and grows) and therefrom must be concluded that float increase must be worth more than profit. It is central to valuing BRK:s look-through profits, which is the reason I bring it up again.

 

Under question 27 on the link below is written a nice explanation:

 

https://buffett.cnbc.com/video/1998/05/04/morning-session---1998-berkshire-hathaway-annual-meeting.html

 

"27. Berkshire insurance float has a negative cost

WARREN BUFFETT: Zone 8.

AUDIENCE MEMBER: My name is Hutch Vernon. I’m from Baltimore, Maryland.

My question has to do with float. You said in the annual, and you’ve said in the past, that float has had a greater value to Berkshire than an equal amount of equity.

I wondered if you could clarify that statement. Is that because the float has been generated at such a low cost relative to an imputed cost for equity, or is there something else behind that statement?

 

WARREN BUFFETT: No, it’s because the float, which is now, we’ll say, 7 billion, comes to us at a negative cost. We would not make that statement if our float was costing us a couple percent a year, even though float would then be desirable. Highly desirable.

But our float is even better than that, or it has been, and so it comes to us with a cost of less than zero. It comes to us with a profit attached.

So if we were to replace — if we were to get out of the insurance business and give up the 7 billion of float and replace it with 7 billion of equity, we would have less going for us next year than under the present situation, even though our net worth would appear to be 7 billion higher.

And I have said, if we were to make the decision — if we were offered the opportunity to go out of the insurance business, and that 7 billion liability would — as part of that decision — would evaporate from our balance sheet, so that our equity would go up 7 billion, with no tax implications, we would turn down that proposition.

So obviously we think that 7 billion, which is shown as a liability, when it’s part of a — viewed as part of an insurance business, is not a liability at all in terms of real economic value. And of course, the key is not what the float is today, and not what the cost is today.

The key is what is the float going to be 10 or 15 years from now, and what is the cost going to be 10 or 15 years ago. And, you know, we will work very hard at both increasing the amount of float and keeping the costs down somewhere close to our present level.

That makes it a very attractive business when that can be done. GEICO’s a big part of doing that, but we’ve got other things, other insurance operations, that’ll be important in that, too. And we may have others besides that in the future.

Charlie?

CHARLIE MUNGER: Yeah. If the float keeps growing, that is a wonderful thing indeed."

 

[SwedishValue]

It’s one very important point that you bring up. In the light of float increase and look through earnings, one likely more closely looking at Berkshire the way Munger and Buffett are - and compared with just the reported GAAP earnings the difference is huge. Buffett spoke about the GAAP earnings compared with look through earnings on many occassions as well.

 

Btw, do I know you? Is the Swedish investing universe really this big? :D

[Cigarbutt]

^The topic does often come up and there have been some equations suggested to discount the liability.

I wonder if the discounting exercise applies to all insurers as the underlying principle means a long term capital commitment related to underwriting discipline.

 

Recently reviewed transactions completed in the non-life insurance runoff area (including by Fairfax) over the last few years. Interesting because the acquirer, in these transactions, aims, in a way, to decrease the advantages related to float:

 

1-cost of float

The acquirer aims for a low cost of float instead of a negative cost of float.

 

2 and 3-discounting and growth

The acquirer aims to accelerate the runoff (active management of claims, commutations etc) and to bring the number of claims to zero.

 

Despite the above, it is possible to buy a runoff book of business at book value (assuming "adjustments" to reserves, strong claims management and superior investment ability) and obtain a satisfactory return. So, reserves liabilities at BH deserve a significant discount. However, unless in a last man standing scenario, growth in float should moderate and float to shareholders' equity has been decreasing and stands now at about a third.

 

When looking at other insurers, interesting to compare the investment leverage (float per share) and to multiply that number by a factor related to excess return expected (or absence thereof) related to superior investment ability.

 

The discounting exercise for BH (balancing float assets and reserve liabilities) also helps to evaluate how much cash could be converted to stock investments if stars align. I would say about 30 to 40% of cash float could be used pretty much overnight.

[Swedish_Compounder]

That is the core of the matter. There is no other major insurer like BRK. Therefore, the same methods can not apply. What Buffett says is that their float should not be discounted at all, since it is expected to grow and it comes at a negative cost.

 

He also makes the example that he would not be willing to give up his float for good if receiving the corresponding amount in cash. Thus, the float is worth more than the corresponding cash amount (in Berkshires case).

 

 

 

[IceCreamMan]

Touching back on a previous topic, what's the explanation for Berkshire's stock price rising 5% the day after the buyback threshold was loosened, but declining slightly the day Buffett said in an interview that he recently bought back "a little" stock (for the first time since 2012)?

 

July 18: S&P +0.22%, IAK +1.09%, BRK-B +5.27%

August 30: S&P -0.44%, IAK -0.62%, BRK-B -0.72%

 

Even if the market interpreted the buyback policy change to mean that Berkshire was likely to buy back stock soon, or if the new policy was perceived as more significant than actual buybacks, I would have still thought that confirmation of buybacks from Buffett would mean something, too.

[SwedishValue]

I've seen firsthand stocks blip up when Buffett mentioned he's bought them. Like with Apple. Instant market reaction. Buffett said his buyback remarks under opening hours, and the stock didn't even blip. It blew my mind back then and it still does today.

 

I think the market has a materially different view to me regarding the meaning of the buyback announcement.

 

I guess it's likely we know more saturday, or monday, or in the months to come.

 

As has been pointed out in this thread many times before: Buffett can barely offset the cash that Berkshire brings in by the amount he can spend on buybacks in the public market.

 

Also, an update for the volume traded in October:

 

111 558 790 B-shares were traded, which means that the maximum 25% threshold would give buybacks amounting to 5,7 Billion USD can have been made from B-shares.

8 600 A-shares were traded, which means that the maximum 25% threshold would give buybacks amounting to 2/3 Billion USD can have been made from A-shares.

 

All in all, the theoretical max for October is around 6,4 Billion USD.

[nickenumbers]

I've seen firsthand stocks blip up when Buffett mentioned he's bought them. Like with Apple. Instant market reaction. Buffett said his buyback remarks under opening hours, and the stock didn't even blip. It blew my mind back then and it still does today.

 

I think the market has a materially different view to me regarding the meaning of the buyback announcement.

 

I guess it's likely we know more saturday, or monday, or in the months to come.

 

As has been pointed out in this thread many times before: Buffett can barely offset the cash that Berkshire brings in by the amount he can spend on buybacks in the public market.

 

Also, an update for the volume traded in October:

 

111 558 790 B-shares were traded, which means that the maximum 25% threshold would give buybacks amounting to 5,7 Billion USD can have been made from B-shares.

8 600 A-shares were traded, which means that the maximum 25% threshold would give buybacks amounting to 2/3 Billion USD can have been made from A-shares.

 

All in all, the theoretical max for October is around 6,4 Billion USD.

 

SwedishValue,

 

I follow all of your points and I appreciate the math regarding the max cash that BRK could buyback [pittance] relative to their GIGANTIC pile of cash.  Your ultimate conclusion is unclear to me?  What does the share buy back, cash reserves and stock price mean to you?

 

Thanks for the clarity!

[SwedishValue]

I have a few conclusions for myself, feel free to disagree.

 

First, Buffett has - on so many occassions - talked about the punchcard approach of investment. It is imperative that if one gets a good idea, you should make it a big one. Thus, I think it’s likely Buffett will use his buyback mandate in a very aggressive manner.

 

Buffett buying back shares means he thinks Berkshire is significantly undervalued. My default position is to always agree with Buffett - especially on matters of valuation and matters of Berkshire. This should be his sweet spot, so Buffett being wrong about the intrinsic value of Berkshire now that he is buying back stock, is not realistic.

 

Buffett buying back up to 25% of the daily trading volume should put some soft floor on the trading level of Berkshire. This also has the implication that the shareprice is likely to converge to intrinsic value faster than it would have without Buffett buying back.

 

Keeping track of possible buyback volume and comparing to reported volume will give a good indication as of how aggressive Buffett has been. All the news articles I’ve read that mention buybacks argue for much lower volumes than I think is likely.

 

All in all, I think it’s a good time to be long Berkshire - also for ”special short-term reasons”.

[Dynamic]

It will certainly be interesting to see the number of shares outstanding at 30 Sep 2018 and those around Fri 26 Oct (printed at the foot of the front page of the 10-Q) especially as BRK.B has spent half of October at around $215-$224 and then dropped to below $210 and as low as about $197.50 during the session on 26 Oct (I topped up my exposure around that level on Friday).

 

On the old faithful metric of Book Value Per Share, the quarter probably ended around $228,500 per A or ~$152.30 per B share, but has since dropped thanks to the assumed stock portfolio retracing a lot of its gains, alongside the market in general. I'd guess ~$197.50 was just below 1.3x BVPS at quarter end, but just above 1.3x BVPS when adjusting for the portfolio decline and typical earnings over 26 days. 1,3x BVPS has typically been a good buy point, though the rare times below 1.25x BVPS (e.g. Jan-Feb 2016) are obviously even better and limit the near term downside risk enormously.

 

My thought is that that ratio of IV to BV has increased a little and this is recognised by Mr Market, and prices significantly below 1.3x BVPS will perhaps be rarer still over future years. BRK purchased at such a price is likely to return inflation+ 6% to 9% compounding with quite a high probability to my mind, and short term downside risk is likely to be limited to about 10% except in the depths of a major bear market, making Berkshire stock at that price a great place to earn a healthy compound return if held long term with a short-term return distribution skewed significantly to the upside, while retaining optionality close to that of cash, allowing me to redeploy my funds at short notice if I happen to find a bargain high conviction opportunity that warrants substantial exposure (and these high conviction ideas might be once in 3-5 year finds so on average I have plenty of time to compound value while I'm waiting). Berkshire also carries, to my mind, very little company risk, because Berkshire has so many diverse earnings streams and autonomously run subsidiaries operating in only modestly correlated areas of the economy, and because it is famously a prudent long-term capital allocator with a strong aversion to permanent loss of capital.

 

While the opportunities around $185-190 in late July were great buy points to me, I suspect that prices around $198-205 now that the buyback rules have changed will be very likely to have seen significant buyback volume in October, though as SwedishValue points out, it might still do little more than offset the cash inflows from operations given the limitations offered by SEC Safe Harbor guidelines.

[SwedishValue]

”It will certainly be interesting to see the number of shares outstanding at 30 Sep 2018 and those around Fri 26 Oct (printed at the foot of the front page of the 10-Q)”

 

The second part of this statement was news to me. Are you saying we will also get to know part of October’s buyback activity? Great news in that case!

[John Hjorth]

SwedishValue,

 

Or close to the end of October 2018. On the front page of the 2018Q2 10-Q the share count for outstanding shares was as per July 26^th 2018.

[SwedishValue]

Thanks. I'm just very confused. This is not the way it works in Sweden and I just somehow assumed they would not have to file anything that would give away what they have been doing in the 4th quarter. Thanks for helping me out.

[John Hjorth]

Admitted, you're not the only one to be confused here, SwedishValue [ : - ) ],

 

Yesterday I was studying the C 2018Q3 10-Q. C reports at that spot outstanding shares at end of 2018Q3.