<THE BORING PORTFOLIO>
Week in Review
A Nearly Perfect Record
By Dale Wettlaufer ([email protected])
Alexandria, VA (Jan. 15, 1999) -- The Boring Port's progress was negative this week, as two major holdings, Cisco Systems (Nasdaq: CSCO) and Berkshire Hathaway (NYSE: BRK) were down about 4.7% and 2.3%, respectively. Our next largest position, Carlisle Companies (NYSE: CSL) lost 5.2% after topping out near its 52-week high last week. Borders Group (NYSE: BGP), our climatologically hampered bookseller, limped in to close the week down 5.2%, within a dollar of its 52-week low. Pentair (NYSE: PNR), closed off 2 1/2% while Andrew Corp. (Nasdaq: ANDW) ruined our near perfectly negative week by gaining 6% in value.
Andrew started off the week with an upgrade from PaineWebber. We don't know what the upgrade was to or from -- that doesn't concern us in the least because sell-side ratings are useless -- but the analyst posited that wireless infrastructure growth in the coming year will be better than expected this year due to decreasing prices for customers and a resulting increase in capacity utilization. Makes sense to us. We still haven't torn Andrew apart, but we like what this management team has done in seriously repurchasing shares and not just putting out press releases on board authorizations for share repurchases. Not to pull an Andy Rooney here, but don't you hate when a company says it's going to buy back shares and never goes through with it? That really burns my butt, personally.
Andrew's valuation, by the way, really doesn't bother us too much right now. We don't see it selling near its intrinsic value and we don't see anything about its basic business that is that unattractive to us. Depending upon other opportunities, we wouldn't have a problem selling it, however. I've attached a spreadsheet showing a simple earnings discount model showing a value for Andrew more than $5 above its current quote. Here are the spreadsheets, in Excel 5.0 format and Excel 7.0 format.
A word on modeling things like this: The discounted earnings model is a reverse-engineering of what the market thinks Andrew can accomplish over time. The exercise in looking at that is not to reflect what we think Andrew can do. The exercise is to measure what assumptions are built into Andrew's price by the market. By constructing such a spreadsheet, we can look at it and judge whether we think the earnings growth assumptions made by the market are correct. "How can this company achieve these specific results?" then becomes the question we then seek to answer. We work backwards on the price of a company we're looking at. This helps us understand what the market is thinking -- we then make the decision as to whether the market is right or wrong.
We've also included an earnings discount model for Cisco Systems in the spreadsheet. This one is less a reverse-engineering of the company and more a discounting model of the free cash flow we think the company can generate. They're large numbers, sure, but remember that this is not so much a hardware company as it is a software company. Much of its manufacturing is outsourced -- it's much more like a pharmaceuticals company or any other intellectual property company than it is like a PC company or something along those lines. Click above for the model on Cisco.
One of Cisco's main competitors in wide area network switching and remote access concentrators, Ascend Communications (Nasdaq: ASND), agreed to hook up with the former Bell Labs, Lucent Technologies (NYSE: LU) in a stock swap that valued Ascend in the mid- to high-$90 range. That's 58 times the 1999 mean earnings estimate, which compares to Cisco's current price at 69 times the mean 1999 estimate. That's no way to value a company, by the way, but just an interested peer analysis heuristic. While Ascend has talked a lot about its MultiGigabit router, the GRF IP Switch, in the past, I haven't heard much about it and they didn't talk about it in their last press release. The conference call, which I listened to, was half about questions that analysts had on Ascend making its numbers with lots of different accruals, unwinding of accruals, and the like. At least that's the impression that some analysts had.
Personally, I'm kind of glad to see these guys out of the market. I think Ascend was way too concerned with analysts' estimates that it didn't run its business with sensible business policies in mind at all times. You don't let the analysts steer you. You steer the analysts. Or you don't even play that game. But you don't push sales into following quarters and book up reserves when you've made your numbers or pull next quarter's sales into this quarter to make your numbers when you're a little behind. Anyway, the market took a little chunk out of Cisco because of this news, but longer-term, Cisco is the gorilla of the WAN companies and it has a lot of things going that answer our primary questions, which are: "Can this company keep finding opportunities to invest its capital such that return on capital stays high and grows?" and "Are the growth rates of Cisco's addressable markets sufficiently large for Cisco to achieve its goals?"
Tool Time: BancBoston Robertson Stephens initiated coverage of Pentair this week. Industrial growth analyst Michael Mach "1" said he likes these companies because of their growth and their strong franchise positions. "Our objective is to recommend investments that over time significantly outperform representative market indices," said Mach. "The defining characteristics of the companies we will follow in this group are double digit long-term growth potential supported by leadership market positions, strong free cash flow, quality management dedicated to shareholder value creation, identifiable catalysts for positive stock price change; and meaningful share price appreciation potential in the context of limited downside risk,'' he concluded. We'll take it. We've come to like Pentair. It kind of reminds us of our buddy, Carlisle Companies. No word from them this week. We're eagerly looking forward to their conference call, which is always very informative.
That pretty much does it on news. I did want to include something we were talking about in a Berkshire discussion group earlier today. This is a bunch of shareholders that get together to joke around, talk about whatever, and occasionally talk about our company. Only occasionally. The fun often gets in the way. In this discussion, Tex is talking about the cost of float. I respond on that and then talk a little bit about the way we've valued Berkshire that you might find useful. I'll have some short concluding remarks after this:
"For simplicity, I assume that underwriting gains or losses are zero. (To the extent that they are positive, WEB seems to treat them as negative costs of float, and vice-versa.) The stated Geico business strategy seems to be to reduce or eliminate underwriting profits in favor of increasing float growth, but not to the point of taking on over-risky business. Hence, I assume zero [cost of float] for the future...."
I agree with Tex. At one time, I looked at underwriting gains as negative cost of float. I now treat years in which there is an underwriting gain as having the following components:
1. Cost of float is zero.
2. Underwriting profit gets moved into return on capital.
3. Further look-through earnings or investment results go into return on capital.
If you're discounting a series of years, it doesn't matter to the present value of the business how you look at it, though. You could move the underwriting profit into cost of capital and just treat that portion of the capital as having a negative cost. But you must make sure in that case not to enter the underwriting profit into return on capital, because you've already counted it.
Either way, say for simplicity's sake that the only capital invested in the business is insurance float. If you treat cost of capital and return on capital in the following ways, your net present value doesn't change:
1. $100 million in float.
2. $2 million after-tax underwriting gain
3. $6 million after-tax investment income, look-through earnings, or equity return. Whatever have you.
Total return on capital is $8 million, or 8% after tax and the spread is 8%.
The value-added is $8 million -- $8 million in return on capital and zero cost of capital.
If you treat it the other way, you get total return on capital of $6 million, or 6% after tax. But your cost of capital is -$2 million after tax. Your return on capital is lower but your cost of capital is lower, as well. The spread between -2% and 6% is eight percentage points. Your capital base doesn't change, so total value added is still 8%.
On another issue...
In attempting to discount the value of Berkshire over a number of years, I haven't assumed that the cost of float will forever be zero. According to our Chairman, there will be years in which big super-cats will occur. In that case, there will be years in which the company will suffer an outflow of resources and years in which return on capital will be greatly diminished or negative.
In trying to account for this, there are years in my discounting of economic value added that are highly negative. In year nine -- and this is totally arbitrary -- the value added is negative $16.7 billion and the net present value of that is negative $8.7 billion. I've sprinkled some super-cats throughout the spreadsheet in an admittedly random fashion. This wasn't done due to analytical laziness, but because catastrophes can happen at random. Sure, insurers have all sorts of models going in their Cray computers, but the bottom line is that it's hard to model acts of God.
On a net present value basis, the largest single year of negative value added is about three times as large, in absolute numbers, as the largest single year of positive value added. Rather than taking a straight-line approach to how much value that can be built over a number of years, we've "stress-tested" the cash flows. Meaning, we've planned for super-cats. I don't have any way of knowing that these cash flows will be right, but they reflect what could be seen as a conservative appraisal of the prospects for Berkshire. Some people build cash flows that reflect the best case scenarios and then use big discount rates to offset the uncertainty that the scenario will happen.
I used to do that myself. The real trick is to try to get the best idea of what can really happen and make the cash flows reflect that. Then you use a discount rate that reflects the real cost of that capital. Reflecting what Mr. Buffett has said about all money spending the same, I've ditched any adjustment to the cost of equity, treating the S&P 500's historical rate of return as the opportunity cost of equity.
Bottom line on that -- I don't know if the exact cash flows of Berkshire will match what I've modeled, but I constructed it using the guidance that we've been provided -- that super catastrophes will happen. Therefore, the general trend in invested capital and cash flows is upward, but there it's not a smooth trend line that you see in almost any annual report. It's pretty jagged. The capital build trend line is much smoother, though there are certainly negative years. The cash flow line is much more jagged, however.
On the basis of a cash flow that is nice and smooth without any boo-boos, Berkshire would be much more undervalued than we project it to be. On the basis that super-cats can and will occur, Berkshire is still undervalued by about $16,000 to $19,000 per share. I'm sticking with the latter approach and I'm pretty comfortable with that.
Another friend in this group suggested you check out the 1992 Chairman's letter to shareholders for a very good discussion on valuing the float, which is of paramount importance to valuing Berkshire. In fact, this is a major recurring topic in our Chairman's letters to shareholders, since it's so important. If you have free time this Dr. Martin Luther King Jr. holiday weekend, take some time and read at least one of these letters to shareholders. I think you'll be glad you did if you've never read one.
Have a great weekend. I'm winging off for the next couple of days, so I won't see you on the boards 'til Tuesday (remember that band? Yuck). Our Web boards are down for 24 hours or more on Saturday and Sunday for the techie gnomes to do some work on them, anyway. But stop in when you have the chance. We have a great group on the Web and on AOL.
Have a great weekend.
- Discuss Boring Investing on the Boring Port message board.
- 10/01/98: The New Boring Port Transitions Facts
Change the World... work for the Fool.
|Recent Boring Portfolio Headlines|
|10/30/00||American Power Conversion's Ugly Earnings|
|10/23/00||Cisco's Formidable Challenge|
|10/16/00||Cisco, Apple, and Probabilities|
|10/09/00||Perils and Prospects in Tech|
|10/02/00||Learn From Mistakes|
|Boring Portfolio Archives »|
</THE BORING PORTFOLIO>
Stock Change Bid
ANDW + 1/2 19.94
BRKb +23 2128.00
BGP - 1/16 18.63
CSL - 5/16 49.06
CSCO +5 5/16 101.69
PNR - 5/16 39.31
Day Month Year History
BORING +2.12% -2.38% -2.38% 31.08%
S&P: +2.56% 1.14% 1.14% 106.72%
NASDAQ: +3.14% 7.09% 7.09% 125.58%
Rec'd # Security In At Now Change
6/26/96 225 Cisco Syst 23.96 101.69 324.48%
8/13/96 200 Carlisle C 26.32 49.06 86.37%
2/28/96 400 Borders Gr 11.26 18.63 65.46%
12/31/98 8 Berkshire 2244.00 2128.00 -5.17%
4/14/98 100 Pentair 43.74 39.31 -10.13%
1/21/98 200 Andrew Cor 26.09 19.94 -23.58%
Rec'd # Security In At Value Change
6/26/96 225 Cisco Syst 5389.99 22879.69 $17489.70
8/13/96 200 Carlisle C 5264.99 9812.50 $4547.51
2/28/96 400 Borders Gr 4502.49 7450.00 $2947.51
4/14/98 100 Pentair 4374.25 3931.25 -$443.00
12/31/98 8 Berkshire 17952.00 17024.00 -$928.00
1/21/98 200 Andrew Cor 5218.00 3987.50 -$1230.50
</THE BORING PORTFOLIO>