Keep a Critical Eye
Board: Berkshire Hathaway
Those reading through the proxies of the Heinz transaction and other recent acquisitions – Lubrizol and Burlington Northern for example – will have noticed that a large portion of the proxy documents discussed the included ‘fairness opinions’. We’ll also notice that in each of those cases:
(a) the company being acquired (the target) but not the acquirer (Berkshire) obtained opinions as to fairness;
(b) in each deal either two or three separate fairness opinions were commissioned;
(c) the reported fees for these analyses were quite hefty, typically $3 million or so each
(d) the providers of these opinions all had a very substantial additional financial incentive – far greater than the fee for the opinion itself – to see the deal consummated.
(e) while we might expect some similarity in presentation and certainly in factual data, in these opinions we also see some remarkably identical judgment calls among the preparers for respective targets, leading us to wonder how much c.y.a-inspired collaboration might have been involved versus independent analysis.
Some years back I went through a period where I prepared some fairness opinions -- smaller transactions of course, at much, much more reasonable price points -- and I could never shake the feeling that doing these analyses was a bit easy in some regards, sort of cheating. On the other hand, the level of care, accuracy, and documentation required for this particular type of analysis could be fairly daunting.
The easy part: Besides paying well relative to other analytical work, the premise was pretty simple. With these opinions the preparer doesn’t have to arrive at some specific value as a result of several types of analysis, or even narrow range of value. We just have to demonstrate that the price being paid in the transaction is ‘fair’ to our constituents – that the purchase price is higher than the various results that we’ve come up with. Putting this in Berkshire terms for a moment, we all know that it’s far easier to make a compelling case that Berkshire is worth ‘at least’ some number than it is to derive and successfully defend some specific number. In fairness opinions, we just need to do the former. We also don’t need to make any assurance that some other potential buyer might be willing to pay more, or might see more value in the target – strategic or synergistic, for example – just that the particular deal on the table is ‘fair’.
On the other hand, the work that is done requires an extraordinarily high level of care. The reason for these at all is to have pre-established the basis for a defense against possible shareholder lawsuits, meaning every factor or assumption that is used has to be supportable. In those published summaries in the proxies we see some annoyingly vague verbiage, “in our opinion and based on our experience” language, but in the supporting documentation – and most regrettably that detail is not included for us in the proxies – we can be sure there is a foundation, a supportable thesis, for each factor, number or assumption.
That said, while this kind of rigor can be tiresome in large doses, it’s probably healthy to subject ourselves to this kind of thought discipline to from time to time. It’s a matter of challenging ourselves as devil’s advocates at each step and for each assumption, in anticipation that unsatisfied shareholders’ experts might do the same. It’s not a bad exercise to have to examine analytical thought processes entirely in quantitative terms from time to time.
This work is ‘one strike and you’re out’ stuff. If the analysis ever does see a courtroom and if for whatever reason the court finds the work unworthy of an ‘expert’ in the field, the not-competent label sticks and can forever be used to discredit the preparer (permanently for the individual, not necessarily the entire firm, though the stakes are very high there also). The folks on the other side of the case would consider it a good day’s work if they if they could land that permanent knock-out punch to the issuer of the opinion. OK – we’re thinking I’ve beat this horse enough.
Getting back to the proxies -- while there is a lot of repetition in these multiple opinions, and they can be suspiciously similar, the content can be interesting even in their frustratingly summarized form. Every so often someone comes up with some refreshing point of view, some analysis that we might not have considered before. In any case, in reading these we’ll usually see a lot of bases being covered. For example:
==> We’ll see a comparison of the pricing of this acquisition to other transactions of that scope.
==> We’ll also always see a comparison to the current market pricing of similar publicly traded companies – applying the market prices to both trailing twelve month historical earnings and forward year projections.
==> There will be a discounted cash flow analysis (more on that in a bit).
==> While it is pretty basic, we’ll get a comparison of the purchase price to the target’s recent stock prices, its 52-week and all-time highs, etc.
==> Sometimes – for example with Citi’s analysis for Lubrizol -- we see an ‘LBO’ analysis. This is a good gut-check and as a matter of principal I like to see that this has at least been considered -- as selling shareholders we should at least have an idea of what a fully leveraged buyer could stretch to pay, and what assets might be monetized
==> While they aren’t showing it in the proxy summary, we can be fairly sure the analyst at least gave some thought to a liquidation analysis, as a floor.
I’ve seen fairness opinions that have included a dozen distinct analytical approaches to value, the idea being to make sure we haven’t overlooked some aspect of less-than-obvious value. Reports like that can be educational when we find them.
In any case the preparer doesn’t want to be blind-sided by a disgruntled shareholder who might have some legitimate view on value that the fairness opinion didn’t consider. At a minimum, the three basic analytical ‘food groups’ have to be addressed:
(1) market based comparative valuation (in each of these three transactions we see both (a) market multiples of peer companies,(b) past M&A pricing for somewhat similar deals);
(2) an estimate of intrinsic value of the business’ cash generation; and
(3)an appraisal of its net assets – a Graham-type balance sheet review.
If the price paid is fair based on each of these general approaches and we don’t see any other elements that would cause us to think the proposed price is too low, we get the ‘fair’ certification.
A Berkshire Fairness Opinion?
Stepping back a moment, there is no legal requirement that managements get a fairness opinion. Some companies never get them. I don’t recall ever seeing Berkshire paying for one on its shareholders’ (our) behalf, and probably for good reason.
First of all, we can’t picture Buffett (or Charlie) voluntarily paying a few $mill to have one of Tracy Britt’s former classmates do a couple of weeks’ of analysis to assure us that Buffett is getting us a good deal, on either a sale or a purchase. Further, it’s not easy to imagine a disgruntled shareholder dragging Buffett into court because they thought Berkshire screwed up (even if a transaction did turn out to be an eventual mistake). We might send the shareholder to Charlie for some suggestions on what that unhappy person could do with his shares. Even if a complaint got to court somehow, we can’t see any ‘expert’ successfully taking on Buffett over the value of a deal -- or anyone even wanting to try.
And perhaps there’s something else. To the extent that Berkshire shares might be involved in the transaction, the preparer would need to obtain some best-effort future-year projections of cash flow from management, and would also have to develop a discount rate for Berkshire – items we’d all be interested in seeing for sure, but that we know management would not want to have to address publicly (or implicitly endorse).
I’m sure someone will correct me if there has ever been an instance that I might have missed, but I don’t envision Berkshire directly commissioning a fairness opinion involving itself, at least not on Buffett’s watch.
However for other company managements, the mere mortals, it’s not so simple. It’s not a bad idea to get an independent point of view on an acquisition – assuming it’s both by a competent source, and that it really is independent (more on that shortly).
Fairness opinions really got their start in the mid-80’s. Back then there were some court cases – notably one decided by the Delaware state supreme court - that found managements negligent for selling out to acquirers, even at apparently compelling premiums to market prices, without at least getting a third-party, objective opinion. In that landmark Delaware case, the acquisition was at a 50% premium to market, and the court still found management negligent.
We’ll remember that this was back in the Drexel and KKR hey-days, before many managements (and for that matter their investing public) started thinking out-of-the box in terms of maximizing or monetizing underperforming assets, and the market really didn’t always fully appreciate the value of what companies might own. Based on what was happening in the M&A markets, particularly with LBO’s (and not to mention often-conflicted management buyouts) the courts had legitimate concerns that minority shareholders were getting unbiased and competent information and analysis for proxy voting.
With that legal backdrop, any managements not bothering to get an outside opinion for a transaction became sitting ducks. And so a niche financial business was born -- a lucrative one at that. As we see in these recent Berkshire deals, quite a few of Wall Street advisory firms perform these services. As with most financial services, we pay mightily for opinions from the big names, the Goldmans, the Citi’s and such.
A side note: while we won’t see them participating in Berkshire-sized megadeals, or charging outsized Wall St ‘bulge bracket’ fees, easily the largest provider of fairness opinions nation-wide is the firm Houlihan Lokey. They are pretty much the 600-pound gorilla in this analytical niche for the vast mid-market of American business. Their customers can’t afford and aren’t hiring the Goldman’s for transaction services. Numerous boutique firms are also in the segment, often specialists. As one active practitioner once noted, any opinion is only as good as its expert on the witness stand, where company size or pedigree doesn’t carry weight.
We’ll talk about the Wall St. majors’ apparent conflicts of interests in a moment. Whether it’s Houlihan or one of the many specialist firms, there is at least some obvious advantage to engaging a firm just for the fairness opinion, a preparer that’s not also at the client’s trough for more lucrative services – especially related to that same deal.
Three recent Berkshire-related proxies.
Heinz got its three(!) fairness opinions – that the deal was in fact fair to their existing shareholders - from its deal advisers Moelis, Centerview, and BofA/MerrillLynch. Combined deal fees for the three advisers totaled $57 million, $47 million of which was contingent on the deal actually closing. A total of $10 million was paid to the three upon delivery of their fairness opinions of the transaction. Guess what. All three assured Heinz shareholders that they were getting a fair shake.
Lubrizol shareholders got their nod from Lubrizol advisers Citi and Evercore. These two split $52 million of fees, $46 million of which was contingent on the transaction going through. Earlier on, though, as the proxy was being assembled they got $3 million each for delivering their fairness opinions to Lubrizol shareholders, endorsing the deal. By the way, Citi included an LBO analysis (mentioned above) among their several approaches to analysis -- when I noticed that I mentally gave them a bit of extra credit for at least looking at that.
BNSF used Goldman Sachs and Evercore. There was no breakout of a fairness opinion fee, but if they did sign off on ‘fairness’ and the deal actually closed, which it did, Goldman was to get $35 million and Evercore $11.5 million – again, contingent on closing. While perhaps only superficially, Evercore at least made an attempt to assure BNSF shareholders not only the price was adequate, but also that Berkshire shares, if they elected to take shares, were fair currency.
All of the firms freely explained their conflicts of interests. Goldman’s was particularly out there. They reminded BNSF shareholders, in case anyone might not have known, that they were particularly beholden to the acquirer for saving their bacon with a $5 billion preferred stock infusion right when they needed it most. And yes, Berkshire was an important client. And yes, Burlington shareholders could take comfort that the Berkshire offer was fair to them (BNSF shareholders).
Here’s the thing. If there are undisclosed conflicts, the fairness opinion can lose its validity and can get summarily tossed in a dispute, and the provider can be subjected to its own litigation. However if potential conflicts are simply disclosed, no problem there. Everyone has been warned, and can draw their own conclusions. Besides, even if there might be an appearance of conflict, we can compensate for that with volume, right? How could three (potentially conflicted) opinions that reach the same conclusion all be wrong? (yes, I’m being facetious)
Again, there is nothing illegal or patently improper here. There is no requirement for the opinion, and we might conclude that something is sometimes better than nothing (we know it is for the targets’ management, despite the cost). We can also expect that the targets are probably ‘paying for it anyway’ in that it’s likely part of the overall deal package, and advisory fees wouldn’t be discounted by $3 million or whatever if we deleted that option.
A third party opinion likely really would be incremental cost, perhaps not $multi-millions, but something. (But we might ask, why might management invite in a potential nay-sayer, even if it was ultimately in their best shareholders’ interests? Where would they be then if they got a negative answer?) In any of the three deals, however, a truly non-conflicted opinion might have been a refreshing indication of good intent. But I’m digressing.
Discounted cash flows
With DCF’s there are of course two key elements: financial projections and discount rates. While skeptics might be wary of company management projections – and BNSF and particularly Heinz were reluctant to share anything like that publicly – we should also remember that the information provided in proxies like these is not taken (or presented) lightly.
On the contrary, we see disclosures in these that are enlightenments to even the principals in the deals. Remember that Buffett mentioned that he first learned the real extent of Citi’s and Sokol’s activities in Lubrizol as he was perusing the proxy. Participants were more diligent about accurate disclosure, and apparently fearful of penalties, than they were in communication with their own boss.
Buffett has long explained that his mental model for ‘intrinsic value’ includes some evaluation of the probable discounted cash flows he expects from the business, and has even commented that he incorporates information like prevailing interest rates into his private views about discount rates.
To the first point, we know what he thinks about ‘management projections’, but he must have some inkling of his own expectations from businesses. If nothing else, through his vast personal experience and his own private information network of diverse subsidiaries, we can be reasonably confident he can size up prospects.
Shareholders of Berkshire’s acquisition targets, however -- the rest of us looking in from the outside, more likely need some help. We can see what managements say they are thinking and make our own assessments as to how we use that information. If we have a bias, it is to be sure we aren’t being presented with too-conservative numbers so that we are steered into relinquishing our shares too cheaply.
BNSF (alone) took the rather clever approach of presenting several forecasts, having captions like “Deeper Recession”, “Recovery Case” and “No Recovery”. Apparently not wanting to sand-bag BNSF’s shareholders’ ‘fair’ price expectations, analysts smartly used “Recovery Case”. BNSF meanwhile side-stepped future criticisms if results didn’t quite pan out.
Heinz noted that it had prepared two forecasts for internal use – one ‘as-is’ and the other including possible ‘bolt-on’ acquisitions, and that only due to the requirements of the impending transaction had shared the ‘as-is’ version with us in the proxy.
One quick check that any outside analyst using management projections should hopefully do is request past years’ projections – typically those that had been provided to the board for then-future periods—and compare those old projections to subsequent results for the projection period. We can get a pretty quick sense of managements handle on these things, and its bias in forecasting – whether overly optimistic or habitually understated.
In any case, the purpose is to get existing shareholders, who wouldn’t otherwise be privy to this inside thinking, enough information to provide an informed vote.
When it comes to discount rates, we have reasonable confidence that Buffett can well make his own judgment of industry and company expectations and the appropriate factor for probabilities of attainment, moat safety or risks, and such. For targets’ shareholders, we can see what those highly compensated ‘independent experts’ have come up with, and make our own judgment. As shareholders asked to sign off on the sale, we here again first off want to check that the values we are being offered aren’t sandbagged – that we aren’t be shown too high a discount just to sell us on a lower sales price.
My own feeling is that if anything, minority shareholders are conditioned to apply, and accept, equity hurdle rates that are too low (the flip side being a tendency to overvalue and perhaps overpay). For our purposes here though, the objective is to make sure we are being adequately compensated, so we don’t want to be biased towards an overly conservative (high) discount rate.
So what did our experts come up with? For BNSF, Evercore said ‘8% to 10%’ ; Goldman said ‘7% to 11%’ – a broader band with the same midpoint. For Lubrizol shareholders, Evercore came up with a firm ‘11.5%’; Citi gave us a range of ‘10.5% to 12.5%’ , which of course gets us thinking ‘11.5% mid-point’. So far, the experts agree most remarkably with each other in their respective transactions. Great minds must think alike, etc.
Heinz shareholders – who again paid $4 million apiece for two of their expert opinions and $2 million for the third, were told exactly the same thing by all three: a range of ‘5.5% to 7.5%’. We might be excused here for suspecting that these three high-end fairness opinions might not have been prepared and submitted completely independently. At least from the selling shareholders’ perspective, the numbers were thankfully not manipulated to show too-high rates (resulting in lower value). But still, this is too convenient, too blatant. If I had been a shareholder and had had a forum, I might have asked.
Similarly, we can look at the provided 'market-based' analyses within the fairness opinions, where one firm picked 10 or so 'peers' - somewhat similar publicly traded companies - to get some idea of the targets' comparable value, and guess what -- the other opinion provider happened to select the same group, exactly, for its analysis. But as readers of these reports, we already get the point.
So, do we need standards? Legislation? Reform? Those questions get kicked around from time to time by legislators and regulators. Should they be required? And will we ever see Berkshire commission a fairness opinion? We can expect that they would be a poor use of time and money for Berkshire, at least under current management, and might even work against us. We like, and benefit from, Berkshire’s holding its cards close. The danger of legislation or regulation is that we get unintended consequences. Further, we should never think that these analyses are ever a substitute for independent thinking – good judgment – by shareholders.
We don’t want to give these reports any more weight than they deserve. Hopefully though, enquiring minds among shareholders and their board members can gradually get us something more meaningful. If we can contain the outsize fees and hire with an eye towards independence, these really can – and often do – give minority shareholders some useful information on with to base their votes. The irony is that, while we only touched on but really haven’t fully explored this tangent here, it would seems the in some respects the more ‘mid-market’ the deal, the better, more useful, and less conflicted the information.
If there is meaningful upgrade, or some kind of shake-up in all this, maybe it will come from the same source that got us these analyses to begin with – unhappy shareholders and (regrettably) their capable attorneys. Maybe we can collectively introduce some common sense to these larger deals. In the meantime, we’ll probably want to take these reports we are provided with as what they are. We can continue to look for the valuable nuggets – there almost always are some, intentional or not – but also apply a critical eye.