Leucadia-Jeffries Merger Loose Ends

So after posting my initial look at the merger, a couple of points have been raised.  Or I should say one was raised (in the comments section in the previous post) and that lead to another point.

One is the issue of dilution.  I said that this merger would have a 4.7% or so dilutive effect at the prices on the LUK presentation.  This is correct on a pro-forma, post-merger stated book value basis. 

But someone pointed out correctly that the book value of JEF in that calculation uses LUK's acquisition price for JEF and not the book value of JEF; the post-merger book value of LUK would include the premium to book value that LUK pays for JEF (which would end up as goodwill).

Is this fair?  I will get to that in a moment.  I don't think it's a bad analysis because we still look at Berkshire Hathaway, for example, using book value per share even after the Burlington Northern acquisition (we don't adjust the BNI value down to the pre-merger book value of BNI).   Book value is only a rough guide and there are other models to look at BRK, but still, BPS is an important measure for BRK's value; important enough for Buffett to use 1.1x book as a buyback level.

But anyway, when we look at financials, we do tend to look at if it is dilutive or accretive to book, and we do often look at stated book value of the respective firms.

"Real" Dilution

OK, so let's recalculate the dilution to LUK shareholders of this deal excluding the goodwill that may arise from it.  In order to do that, I will use JEF's current book value instead of LUK's purchase price of JEF, add it to LUK's current book value and get a post-merger BPS.

Here is the page from the presentation.  If you can't see it, it's page 32 of the merger presentation available at LUK's website.

 

So instead of using the LUK share price times 0.81 x JEF shares outstanding to calculate JEF's adjusted book value, let's just use their stated book value.

According to JEF's latest 10Q, JEF's adjusted shareholders' equity is $3,515 million.  So let's use that instead of the $3,782.1 million adjusted JEF equity value.  To calculate this quickly, all we need to do is adjust the combined book value for all of LUK down by the difference which is $267 million ($3,782 mn - $3,515).

The adjusted combined book value for LUK in the above table is $9,325 million.  Deduct $267 million from that and you get $9,058 million.  Divide that by the adjusted LUK shares outstanding after the merger of 377.7 million and you get a post deal book value per share of $23.98/share.

According to the above table, LUK's current BPS (less Crimson) is $25.91/share.  That means that excluding the goodwill from the deal, dilution to LUK shareholders is close to 7.5%.

So that looks like a lot of dilution!

Obviously, if you think JEF is worth only book value, then you would be pretty unhappy with this deal.  Not good.  

But wait a second.  All the time we have been valuing LUK, we have marked the LUK holdings to market.  When JEF was trading far above book value, we didn't adjust book downward to JEF's stated book value.  We marked the position to market.

So in that sense, we don't need to force the value of JEF to book value just because they own all of it.  Why mark to market when partially owned and then force a mark to book value (and exclude acquisition goodwill) when wholly owned?  That is not consistent. 

Of course, recently, JEF has been trading lower and LUK's deal has pushed the stock price up so one can argue that appreciation in price is artificial.    But let's take a closer look at this and see what JEF might be worth (or at least what maybe the LUK folks feel it's worth).

JEF Value

When LUK first bought into JEF in April of 2008, the deal was priced on April 18. The closing price of JEF on that day was $14.98/share.  As of the end of the most recent quarter at the time, JEF's book value per share was $13.03 and the adjusted BPS was $12.10 (adjusted for RSU's).  

So that was done at 1.15x BPS and 1.24x adjusted BPS.  Granted, at the time, LUK's share price was $53.36/share against a BPS at 2007 year-end of $25.03/share.  So they issued stock at 2x book to buy something at 1.2x book.  Nice trade.

But LUK has also bought stock in the open market over time.  They have said previously that JEF is a good buy at close to book value.

So JEF has been on the LUK books since 2008.

I just jotted down the BPS, adjusted BPS and stock price of JEF since 2009 to see how JEF has been valued in the market.  I figured 2008/2009 stock price would be depressed due to the financial crisis and not really that representative and anything before that of course may be even more unrealistic as financials were valued pretty high pre-crisis.

Anyway here is the data:



JEF Valuation



Common BPS is the GAAP BPS that most people see.   This is the figure people see when they think the LUK deal was done at below book value.  The adjusted BPS is adjusted for RSU's and is more accurate since it is fully diluted for RSU outstanding.


Anyway, since the first quarter of 2009 (the panic low during the crisis), JEF has traded at an average of 1.56x adjusted book value per share.  1.56x is not a bad benchmark as it is mostly a post-crisis valuation and it includes the MF Global contagion/JEF panic in late 2011.

If you exclude the early 2009 panic low and the MF Global panic and only include the quarters 2Q2009 through 2Q2011 (labeled "average*" in the above table), JEF traded at 1.96x adjusted book value per share.

So that is sort of the market evaluation of the JEF business model post-crisis excluding the effects of the Euro-meltdown/MF Global panic. 

JEF has not recovered from that, and the market is now undergoing a fiscal cliff panic.  So excluding these effects, JEF may be reasonably be valued at 1.5-2.0x BPS.

Going Back to the Dilution Issue
So the first point would be, if we didn't adjust downwards the value of JEF to it's stated book (when it traded above book) and we calculated LUK's book value using the mark-to-market value of JEF when it was partially owned, then it doesn't make sense that we do so now post merger just because this "goodwill" is artificial. 

The part that is artificial is that LUK's acquisition did push up the price of JEF; it doesn't reflect pre-merger mark-to-market of JEF itself.   But it didn't do so beyond the previous range of where JEF has traded in the recent past.

With LUK trading at around $21/share now, that makes the deal worth $16.35 per JEF share. Since the adjusted BPS of JEF is $15.63/share, that's a 4.6% premium, not much.   Even at the presentation value for JEF of $17/share (when LUK was at $21.80), that's a 9% premium, or a value of 1.09x book value for JEF; not an unreasonable valuation at all given it's recent trading history.

You can argue that this very deal will reduce or eliminate the liquidity concern that arose about JEF after the collapse of MF Global; there is no reason why after this deal that JEF should be worth less than book.  As an independent, there was always a concern of an MF Global-type run triggered by a European collapse or some other event.

So I would be inclined to say that the LUK presentation post-merger valuation (including goodwill) is fine for looking at dilution.  I think it's important for shareholders to understand all of this, though, and realize that there is another way too look at it (7.5% dilution).

And By the Way
Also, there is a circular aspect to this dilution calculation and what LUK is trading at now versus a post-merger BPS.

Since the deal is not fixed in terms of price (no price floor/collar or anything like that), we don't really know what the dilution is going to be.

Calculating the above LUK presentation dilution (using LUK acquisition price as JEF adjusted equity), the dilution changes as follows at various price levels of LUK:

LUK Price           post-deal BPS        dilution
$25                       $26.21                   1.1% (accretive)
$24                       $25.73                   -0.7%
$23                       $25.25                   -2.6%
$21.80                  $24.68                   -4.8%  (price on presentation)
$21                       $24.30                   -6.2%
$20.32                  $23.98                   -7.5%  (price at which deal value is equal to JEF adjusted BPS)

There would be no further dilution under $20.32/LUK share because if the deal price was worth less than adjusted JEF BPS, there would be a bargain purchase gain, and JEF will be booked at the original JEF book value.

There is obviously less dilution the higher the LUK price upon the closing of the deal because the JEF adjusted equity value will include "goodwill"; it will be booked at a higher value on the balance sheet in proportion to how high the LUK stock price is.

With LUK shares trading now at $21/share, the immediate dilution to LUK shareholders (including acquisition goodwill) would be -6.2%, more than the -4.8% calculated according to the presentation (due to the lower LUK price since then).

Post-Deal BPS and LUK Valuation
One other thing is that since the post-deal BPS is dependent on the price of LUK at the closing of the deal, the post-deal BPS on the LUK presentation is not current since the stock price has changed.

I said that the post-deal LUK BPS is $24.69, and if you add back Crimson Wine, then the value is $25.50.   Against that, it looks like LUK, which closed at around $21.00 today is trading at a 17.7% discount to post-deal BPS.

But this post-deal BPS is only good with LUK priced at $21.80.

With LUK trading at $21.00/share, the post-deal BPS is actually $24.30, and with Crimson added back in, that's $25.11.

So the actual discount is 16.4%, not 17.7%.  Not that big a difference.  It's still pretty cheap.

Anyway, here is a table that shows how the discount changes according to LUK's price:

LUK                          post-deal                     
price                          LUK BPS                  discount
25                              27.02                           -7.5%
24                              26.54                           -9.6%
23                              26.06                         -11.7%
21.80                         25.50                         -14.5%
21                              25.11                         -16.4%
20.3                           24.79                         -18.1% 

The above post-deal BPS includes Crimson Wine to make it comparable to the current price.
Below around $20.30/LUK share, JEF would be valued below book, so there would be a floor there because JEF would be recorded at their own book value with the difference recorded as a bargain purchase gain.

Of course the other way to look at the post-deal BPS is to use JEF's current book value instead of LUK's purchase price.  This would not change because of changes in the LUK stock price.

From the above calculation, this would be around $24.00/share.  So even with JEF booked at the old book value and not including goodwill that may arise, LUK's post-deal BPS would be $24.00/share.  Adding back Crimson Wine's $0.81/share value and you get $24.81/share.  With LUK trading at $21/share, that's a 15.4% discount.  Not bad.

LUK is Still Cheap
So even with the above adjustments and high dilution due to a lower LUK stock price, LUK stock is still pretty cheap, and even if with no acquisition goodwill, LUK is trading at a 15% discount to the post-deal BPS. 

So Why'd They Do It?
If you asked them why they did this deal despite the dilution, I bet they would say that they think that the deal is accretive to LUK on an intrinsic value basis.  They will tell you that they are not too concerned with conventional accounting, GAAP book values and things like that.

They did say that they really like JEF at close to book value, which means they think it's worth much more.  They were comfortable owning this at 1.5x-2.0x book for much of the post-crisis period.

If you think JEF is worth 1.5x book, for example, then this deal may not be dilutive at all.  The trick is figuring out what LUK is worth, of course.

Dilution Based on Instrinsic Value
Just for fun, I will fill in some of the above tables and see what the dilution would be if JEF was in fact worth 1.5x book value.  I know some of you will think I am reaching here and trying too hard to make this deal look reasonable.  But that's OK.  I am just looking at this from different angles and I'm not trying to argue one way or the other.

Using the LUK presentation table, let's just wave a magic wand and change the number in the adjusted JEF equity value to 1.5x JEF adjusted book value.  Adjusted book value of JEF was $3,515 million at 3Q-end, so 1.5x that is $5,273 million.

Now the post-merger LUK BPS comes to $28.63/share.  The current LUK BPS (excluding Crimson) is $25.91/share, so now the deal is suddenly 11% accretive.

And then adding back the value of Crimson, LUK BPS would be $29.44, so that would make LUK at $21/share trading at a 29% discount!

OK.  So there are problems here. I used an intrinsic value of JEF against the book value of LUK.  To be totally fair, you have to look at intrinsic value versus intrinsic value.  But for me, since I usually value LUK at book value (adjusted for some things), it is not too far off.

This is just an illustration of why this deal is not as simple as "it's 8% dilutive so it's horrible!".  Technically it's dilutive, but there is more to the story than that.  And that's what I tried to illustrate with this example.  I don't mean to say that this deal is 11% accretive, or that LUK is now trading at a 30% discount.

I just point out that it can be seen this way if one thinks JEF is reasonable worth 1.5x book value.  And that's not a stretch at all, again, if you think that this deal will reduce the risk of an MF Global-type run, or someone Egan-Jonesing them again.

Of course, if you think JEF is worth 1.5x book, then JEF has been a great value in the past year or so.  Well, maybe that's why LUK is buying them out.


Can We Really Get to 1.5x Book for JEF? 
One last thought.  I think the smaller investment banks do tend to trade at higher multiples than the large ones.  Recent history of JEF trading when not being Egan-Jones-ed tends to support higher valuation for a company like JEF.

Many of us make fun of investment bankers all the time, but I do like to read merger proxies as they do tend to have valuation comps that can be interesting (and silly at times too).  So we may get more insight into valuation for JEF.

But we may be able to get 1.5x valuation on our own.  We know that the LUK folks want to get 15% pretax return (or much better).  So let's use a pretax 15% return.   Since I already did the work, we can use book value growth plus dividends as a proxy for long term return-on-equity (it will serve as a sort of comprehensive return on equity over time)

From the other post, here is the growth in book (plus dividends) of JEF over time (annualized):

                            Growth in BPS+DVD            Pretax Comprehensive Income
1996 - 2011         +15.4%                                 +25.7%
1999 - 2011         +13.9%                                 +23.2%
2007 - 2011           +5.6%                                 +  9.3%

So since 1996, JEF grew book value including dividends at a rate of +15.4%.  On a pretax basis (using a 40% tax rate), that's +25.7%/year.

From the 1999 peak, they earned 23.2%/year pretax.

If they need 15% pretax returns, JEF can be worth 1.5x book (at 1.5x book, this 23.2% pretax return turns into 15%).

Now this too is not so simple.  Yes, the previous ten to twelve years has not been the greatest time for financials but there is no guarantee that the next ten or twenty will be as good for JEF.

This, again, is just to get my arms around a 1.5x valuation.  Is it reasonable or not?

I don't do this to convince anyone either way; I just present the facts and analysis as food for thought.

Conclusion
So I fine tuned some stuff from my other post here and there were some adjustment/changes that had to be made, but I don't think the conclusion changes much.

The deal is dilutive looking at it conventionally. But who said these guys are conventional?

For me what's important is that I do like the people involved on both sides, the deal is not ridiculous (this is not Time Warner / AOL) and in fact might be a great deal depending on how you look at it (1.5x book value scenario), and the LUK stock is currently pretty cheap either way.

The 1.5x book value scenario and analysis is just illustrative and shows that this can be a a totally reasonable deal depending on what you think JEF is worth.  1.5x is just a figure I plucked out of the air as it seemed to be at the lower end of the range of JEF's stock excluding periods of panic (and it's the average since 2009).

I know many people assume investment banking is dead and that's why JEF is trading cheap.  But I tend to disagree with that.  I don't think investment banking is dead at all; I do think it will come back.  And I think JEF was trading cheap mostly for fear of another MF Global-like run.  This is clear from the valuation pattern through 2011; it traded well until MF Global collapsed and a bad (and wrong) report about JEF came out.

Of course, if you think investment banking is dead and the low valuation is the correct valuation, then obviously we will disagree on what we think about LUK going forward, and about this deal.

You can easily make a similar but more moderate argument using 1.1x book, 1.2x book etc.

So my opinion remains the same.  Good deal, but of course it would've been better if it wasn't dilutive, but then again, from a value gained perspective, it may not be as dilutive as it at first appears.

Also, this is not a complicated deal at all.  It's pretty simple.  But there are a lot of numbers involved in the stuff I wrote, so I may have missed something (hopefully nothing big!).  If so, I apologize in advance.

In any case, we all have to do our own work so do your own work and make up your own mind!