Thursday, December 20, 2012

Markel - Alterra Merger



So, Markel (MKL) is buying Alterra (ALTE).   We know Markel is acquisitive and is looking for opportunities.  So here it is.

Of course Markel shareholders might have gotten sticker shock from the stock being down so much.  If the deal is accretive to book value, then why the heck does the stock price have to be down so much?  Markel was trading above book value, and they will buy Alterra for around book value.  But the way the market sees it, at least initially, is that on a sum-of-the-parts basis, nothing really changes on this deal.

Before the deal, the combined market cap of Markel and Alterra was around $6.9 billion, and now the combined market cap is around $6.8 billion, so the book value accretion to Markel (positive for MKL shareholders) is offset by the increasing exposure to reinsurance (negative; reinsurers are all trading below book).

Anyway, the merger presentation (which is always educational) is on the MKL website.

I like Markel, so I tend to like this deal.  It's a good idea.  Buy float for cheap and bump up returns by improving investment returns.

Here are some highlights:

First of all, Markel is a good company with decent historical results:



Markel's strategy is to do well on both underwriting and investing.  Most insurers work on the underwriting but invest mostly in bonds.  

This deal is accretive to book value.  After the deal, the book value per share of MKL goes from $395/share to $424/share.  The stock is now trading at $434/share, around 1.02x post-deal book value per share.  ALTE had $29.57/share in BPS at the end of September 2012 excluding $2/share in unbooked gains on fixed income securities held to maturity (booked at amortized cost).  So this $424/share post-deal is clean and simple (unlike the LUK post-deal BPS that included goodwill from the JEF purchase).

I don't believe MKL would do this deal expecting sub-10% return-on-equity for ALTE going forward, so MKL should be a decent buy at around book value.



Here is one of the keys to this deal.  ALTE's investment portfolio is mostly in cash and fixed income securities with a small allocation to hedge funds (more on that later).


ALTE has increased BPS year-to-date almost 10% (including dividends), which is not bad at all given the low interest rate environment and still weak insurance market.  ALTE's average ROE in the past five years is around 8%, and again, that's in a low interest rate environment and weak insurance/reinsurance market.

So we can see how there can be some pretty good upside from here just from:
  1. Increasing ALTE's allocation out of fixed income into equities
  2. Hardening insurance market
As for the insurance market hardening, here is an interesting slide from ALTE's 3Q2012 presentation:


So, insurance companies are trading really cheap these days because:
  1. Interest rates are low:  Insurance companies are leveraged due to float so their investment valuations are sensitive to interest rates and investment returns.  (At 3x float leverage, an 8% interest rate would mean a 24% pretax return on equity just for holding the float.  In a 3% interest rate environment, that goes down to 9%.  On an after-tax basis, that is the difference between a business valued at 1.5x book and 0.6x book; that's a 60% decline in value. (This assumes 35% tax rate and 10% required return).  Also, rising rates will cause bond prices to decline putting pressure on bond prices.  Some feel that owning insurance companies is owning bubbled up bonds on leverage (well, it is).
  2. Insurance market is soft:  As the chart above shows, the insurance market due to many factors (excess capital, weak economies around the world etc.) has been pretty weak.
  3. Economy is weak:   This overlaps with the other factors, but a weak economy obviously leads to lower activity and less demand for insurance overall.
  4. Weak investment returns:  This too overlaps with the others, but low stock market returns over the past decade has put a lid on valuations of insurance companies that do invest actively in equities.
If you believe the above factors are permanent, then obviously there is no reason to invest in insurance companies even if they are cheap.  But I tend not to think so, even though I do think interest rates will remain low for a long, long time. 

If MKL keeps doing well over time, book value is a great opportunity to get in.

But wait.  Before you go out and buy MKL stock, we have to think about buying ALTE instead.

Merger-Arb
ALTE shareholders will get 0.04315 shares of MKL and $10 in cash for each share of ALTE stock. Here are today's closing prices:

MKL:  $433.95
ALTE:  $28.04

So each ALTE shareholder gets:

0.04315 shares of MKL:  $18.72/ALTE share
Cash:  $10/ALTE share
Total:  $28.72/ALTE share.

The deal is expected to close in the first half of 2013, so let's call that 6 months.  Over that time ALTE will pay $0.32/share in dividends (I haven't seen a comment regarding dividends so I assume they will keep paying as usual.  This may not be the case).

So ALTE holders will receive through the deal closing a total of $29.05/share.

ALTE closed today at $28.04/share so that's a 3.5% discount.  On an annualized basis, that would come to a 7.0% additional return by holding ALTE now and getting MKL later rather than holding MKL now.

But one point is that since this deal is partially cash, you won't have 100% MKL exposure by owning ALTE instead of MKL.  You would own 64% MKL and 36% cash or something like that so if MKL rallies strongly towards the end of the deal, owning MKL outright might outperform.

As a merger arb, not that non-professionals can get 7x leverage at 100 bps funding spread (+50 bps for longs, -50 bps for short), this is what this deal looks like (I used 6x leverage to be conservative in the LUK / JEF post, but I'll use 7x as that is the figure commonly used in equity long/shorts):

If you buy ALTE and short MKL in the correct ratio (0.04315 MKL shares per ALTE share), and you put it on levered 7x and get funded at 100 bps, the above 7% annualized return would become 6% after funding cost and then 7x that would get you to a 42% annualized return.  This is not exact as you aren't long / short an equivalent amount (but I used 100 bps funding spread).

I don't know if that's right, but that's what you get with the above assumptions.  I may have missed something, but as usual, this is not the point of this post so let's move on.

Should you buy MKL or ALTE?
This gets tricky here as there is a cash component.   If you are sure the deal will go through and the deal was all stock, then it's a no-brainer to buy ALTE.  But that's not the case here.

Of course, with interest rates at zero and excess cash sitting in brokerage accounts, you can just lend yourself money and buy more ALTE to get the exposure to MKL.  So lend yourself $10/share of ALTE stock and then go out and buy MKL stock with it and you will have full exposure on the $28 worth of ALTE stock, or lend yourself $14/share of ALTE and buy more ALTE.  This makes sense if you have cash in the account earning nothing anyway; when the deal closes you will get that cash back and it will cost you nothing (as money markets are paying you nothing anyway).

But we don't always have cash lying around so we have to look at things with financing costs.  If you buy on margin, you will probably incur interest expense (which may be high in some accounts) so that may negate the edge you get from the discount.

So let's just look at owning $100 worth of MKL and $100 worth of ALTE.  It is simple math.  You have a discount advantage from owning ALTE, but you don't get 100% MKL upside.  So there is a payoff.   Here is what that looks like:

Owning ALTE versus MKL




The first column, 0.04315 is simply the number of MKL shares per ALTE share you receive on the deal.  Cash per share is the $10/share you get on the deal plus the $0.32 in dividends you would get (assuming ALTE keeps paying for the next six months).    MKL price is simply the different levels of MKL that it may trade at by the time the deal is done.    Return to MKL holder is the return you would get by owning MKL stock.  The bold figure is the current level.

The deal value is the total value of the deal per ALTE share.  The bold figure is the current stock price of ALTE and the last column is the percentage return you would earn by owning ALTE depending on where MKL is at the closing of the deal.

None of these return figures are annualized.

So you notice that owning ALTE is better in most scenarios as long as MKL doesn't rally more than 11-12%.   Up to a stock price of $484, you are better off owning ALTE and above that the MKL shareholder does better.  On the downside, due to the lower 'delta', ALTE is better off in all scenarios.

Of course, this assumes that the deal gets done.  If it doesn't, then ALTE can go back down to the pre-deal price and MKL stock would rally back up so the MKL shareholder would do better.   But I think most would see this as a done deal.

History of ALTE
This deal is very interesting when you think about the history of ALTE.  This is an entity that resulted from a previous merger, but the main entity goes back to 1999 when it was formed by Moore Capital (and other investors) as Max Re Capital.  Moore Capital is the hedge fund firm run by the supertrader Louis Bacon.  You can google him and read all about him.  He has a great track record over time, but hasn't done too well recently.  He is known more as a macro trader than for the strategies represented in the multi-strategy hedge fund portfolio ALTE has invested in.

This deal is similar to Einhorn's Greenlight Re; he set up an insurance company to generate float which the hedge fund would invest.

Max Re's idea was to invest 20-40% of the "float" in hedge funds run by Moore Capital.  It sounded like a good idea at the time.  A multi-strategy portfolio is designed to be uncorrelated to the stock market and other investment strategies within the portfolio.  This was to deliver returns that were uncorrelated and much stabler than the stock market with hopefully similar returns.

A lot of hedge funds are doing or planning on doing these things and the fear is that the insurance company would blow up; they would be too eager to write policies to increase float for the hedge fund to invest and then screw up the underwriting.  That's usually the worry.

In this case, it turns out that the underwriting/insurance business did well, but the investments didn't really pan out too well.   The investments didn't blow up and it did do much better than the stock market with lower volatility, but the returns were not so exciting.

Over the years since 2007, they reduced their allocation to the alternative investments from a range of 20-40% to 20%, to 10-15% and then to below 10% (or some similar reduction over time since 2007).

ALTE Hedge Fund Returns
So this may not be relevant to this deal, but it does show you why this deal makes sense.  And it's also pretty interesting to see what happened here in terms of investments.

Here's the list from the 10K which shows the usual suspect strategies in these typical portfolios:

...and here are the returns for the hedge funds since 2000 through the end of 2011.  This is a multi-strategy portfolio run by Moore Capital.


The annualized returns were:

                       Alternatives         S&P 500
12 years:        +4.7%                   +0.3%     
10 years:        +4.6%                   +2.6%
5 years:          +1.3%                    -0.2%
2011:              -3.6%                      2.1%

If Buffett made his "S&P 500 would beat any group of hedge funds" bet back in 2000, he would have lost in this case because the stock market went nowhere and this hedge fund portfolio did OK.   It has done much better than the stock market over most of these time periods and with much less volatility.

But these funds are supposed to be uncorrelated to the S&P 500 index, and they didnt' even return 5% over time.  That doesn't look too exciting.  Of course, the stock market went nowhere, but that's because it was coming from a very high valuation back in 2000.

The interesting thing is that Max Re started to decrease their exposure to hedge funds back in 2007 when their annualized returns was around 8.4%.  I guess that wasn't good enough either.  (I think people tend to expect 10-15% in these sorts of strategies.)

So, the story is that ALTE (or Max Re) was set up as an insurance company like GLRE (but much earlier) to generate float for Moore Capital to manage.  They probably figured they could earn 10% on their hedge fund investments and do well even with so-so underwriting results, and maybe do very well if they also underwrote well.

It turns out that the investments weren't working out so they ditch the hedge fund strategy and focus on underwriting and invest the float in conservative fixed income.

And then interest rates go down and keep going down and their fixed income strategy suddenly doesn't look too hot.  Going back to hedge funds doesn't look like a good idea as they lost -3.6% in 2011.   Plus the hedge funds only earned 1.3%/year in the past five years.  Not an exciting option.

And then here comes Markel with a more dynamic investment strategy.

ALTE ROE
Anyway, let's take a quick look at ALTE ROE historically for a second.

                ROE
2000       2.7%
2001       0.4%
2002      -1.0%
2003      19.1%
2004      17.0%
2005        0.9%
2006      16.8%
2007      20.4%
2008     -12.3%
2009       17.6%
2010       12.3%
2011         2.3%

Average:  8.0%

So ALTE earned an average ROE of 8.0% since 2000.  The five year average is also around 8.0% so it's pretty consistent.  During the good years of 2003-2007, their ROE was close to 15%.

Of course, with subpar investment results and a pretty weak insurance market, they were still able to earn 8% consistently over time.

Just think what can happen if investment results were improved a little bit.

So how much can investment results improve?

Here is the history of investment leverage at ALTE since 2000:
         
              ALTE Investment leverage
2002            3.1x
2003            3.5x
2004            3.9x
2005            3.6x
2006            3.3x
2007            3.2x
2008            4.2x
2009            3.4x
2010            2.7x
2011            2.8x

Investment leverage has averaged 3.4x since 2002 and 3.3x in the last five years.  It has been trending down due to a soft insurance market, but this may head back up if markets do recover.

Anyway, let's just use 3.0x investment leverage to see what impact investing in equities would have on returns at ALTE.




This may be a little small; if you can't see it, it's from the Alterra earnings supplement on their website for the third quarter of 2012.

It seems like the hedge funds continue to not do too well, returning less than 3.0% year to date, 1.14% in the past year and -0.49% in the past 60 months.  (Only the past 60 months figure is annualized in the above table).  

Cash and fixed maturities are earning 4.7% year to date.

With an investment leverage of 3.0x, a one percentage point increase in the performance of investments would mean a 3.0% improvement in the ROE (since this is Bermuda based, I assume zero taxes).  With a 10% discount rate, a 3.0% improvement in ROE would lead to a 30% increase in the value of ALTE.

Can this be done?  Previously, ALTE had more than 30% of investments invested in hedge funds.  If we take the 30% and invest it in stocks instead, and we assume a 6-8% return in stocks over time, that would lead to an improvement of 4 - 6% over treasuries which yield less than 2% now.  The cash and fixed income portfolio above shows a return of 4.7% year-to-date and annualizes to over 6.0%.  I haven't looked at the portfolio in detail, but I don't think we can bake that into the cake for prospective returns as they are probably largely due to higher bond prices due to lower rates.

So if you can get a 4-6% improvement in investment returns on 30% of the portfolio (assuming much of it is invested in treasuries or low risk securities), that's a 1.2% - 1.8% improvement in portfolio returns, and with 3x leverage, that's 3.6% - 5.4% improvement in ROE.  At a 10% discount rate, that's worth 36% -54% increase in value.

(Yahoo Finance says that even single A-rated corporates yield only 1.5% or so out to five years).

Is Markel's Equity Investing That Much Better Than Hedge Funds?
OK, so after thinking about all of this, the obvious question people will ask is, well, will MKL's equity investment team do much better than ALTE's hedge funds? 

Let's take a quick look at some figures (from the various data above):

                  MKL equity              ALTE 
                  portfolio                    hedge funds
20 years:    +8.1%                       
10 years:    +6.7%                       +4.6%
5 years:      +1.9%                       +1.3%
2011:          +3.8%                       -3.6%

So looking at this, MKL's equity portfolio has done better over time than ALTE's hedge funds;  it has done better in all of the above time periods.

But some may argue that it doesn't look that much better.

Well, to that I will have to remind them that the stock market has been flat recently and that is really not expected to continue forever.

I can see how the stock market, over time, will tend to do OK.  I don't know if that argument can be made with the hedge funds, particularly when they haven't done that well in a period that they would have been expected to do really well.  They are supposed to benefit, usually, from volatility and we had a lot of that in the past decade.

So yes, I would tend to be far more comfortable with an equity portfolio run by MKL than this particular group of hedge funds, especially in light of the fact that institutions seem to be rushing out of equities into alternative investments (and maybe that is why many hedge funds aren't doing too well; too much money chasing the same trades!).

Anyway, I have more to say and I may put together a quick earnings model for ALTE to show how ROE may change with increased investments in equities (but don't hold your breath!  Holiday season coming up and I don't plan on sitting in front of the computer all day!).

But since it's been a while since my last post, I want to hurry up and get this out.








19 comments:

  1. Hi,
    Do you have any information about the taxation of the $10 cash payment? There is some mention in the presentation that this will be a "taxable event" for Alterra shareholders.
    Thanks
    Richard

    ReplyDelete
    Replies
    1. Hi, I don't know anything specific to this deal, but I imagine the $10 portion would count simply as a sale of the stock so you pay capital gains tax on it just as if you sold ALTE stock in an all cash deal, or for that matter if you sold it in the market. On the stock portion of the deal you owe no taxes (even though you would just transfer any gain you had on ALTE to the new MKL position).

      Delete
  2. Thanks for the excellent analysis. I've gone through a similar calculation regarding the ALTE/MKL arbitrage but only penciled in one dividend payment rather than two. That doesn't change the equation very much, however. I've purchased both MKL and ALTE in the days since the merger announcement.

    ReplyDelete
  3. Great stuff.

    kk, I had a question. If you get a chance, shoot me an email.

    ReplyDelete
  4. KK,

    > Also, rising rates will cause bond prices to decline putting pressure on bond prices. Some feel that owning insurance companies is owning bubbled up bonds on leverage (well, it is).

    Does that mean, in general, BV of insurance/re-insurance companies tends to go down during time of high inflation? Also, do they have to mark to market longer term bonds held for maturity, and recognize the gains/losses?

    ReplyDelete
    Replies
    1. Hi,

      Interest rates have been going down pretty much steadily since the early 80s, so I don't think there is much data that would be helpful (in the old days, they didn't mark things to market like they do now).

      But yes, if interest rates start to rise, insurance companies will have losses on their portfolios. But since many of them have shortened duration due to lower rates (and lower opportunity cost for doing so as long rates come down), it shouldn't be that bad.

      The argument usually is that a rising rate environment will happen in a stronger economic recovery and in that case, a hardening insurance market and increasing volume would help offset the headwind of lower bond prices.

      Delete
  5. Great post , what about using options

    ReplyDelete
    Replies
    1. I suppose you can use options here. I haven't looked at anything, though.

      One thing that may be interesting is to see what happened to the implied volatility in the ALTE options. Vols should be much lower now than before since instead of an ALTE option representing 100% of an insurance company, it would represent 65% insurance company and 35% cash or some such; the volatility of ALTE's stock price would be driven by the portion that will be paid in MKL shares upon closing of the deal. $10/share cash would have zero volatility.

      So there might be a volatility arb there (ALTE options against MKL options, if they even exist; I haven't looked). If vol is unchanged in ALTE post deal, you short ALTE vola and go long MKL vola.

      There are different ways to do that. But I don't do those sorts of trades so...

      (actually, ALTE vols can stay high if there is a chance that this deal doesn't get done, so it's not as simple as I describe above...)

      Delete
  6. Any read throughs to GLRE here? I'm thinking it has to be positive, thoughts?

    ReplyDelete
    Replies
    1. Hi, I do like GLRE. They had a problem with underwriting and Einhorn's performance hasn't been amazing lately, but in general I like it. He is not running a multistrategy portfolio like Max Re; Einhorn is running pretty much the whole investment portfolio which I really like (I was never a fan of multistrategy structures).

      Delete
  7. Alterra's reserving seems low to me given recent paid loss developments.

    Alterra has had pretty aggressive growth.

    The CEO had this to say in the 2011 Q4 conference call:

    "We've been a pretty steady releaser of reserves for 6 years in a row and we don't see that changing anytime soon."

    I just had a quick look but:

    Taking the 2003 line of business, paid losses have jumped by $63m last year and $210m the year before yet there is just $42m in reserves left.

    Taking the 2004 line of business, paid losses have jumped by $101m last year and $287m the year before yet there is just $69m in reserves left.

    Am I misreading the P&C triangle?

    ReplyDelete
    Replies
    1. Hi,

      I'm no expert on insurance companies, but I don't think the cumulative paid losses is an indication of under or overreservedness. For that, you can look at the other section, reserves re-estimated and see if their initial reserve estimates have proven to be conservative or not. Deficiency / redundancy tells you that.

      But you have to be careful with that too as oftentimes that may just reflect the insurance cycle; hard markets allow higher prices and higher reserves that may prove redundant over time.

      The best is to compare this deficiency/redundancy with similar insurance companies; that may give you an indication of how well the underwriting is and it might help filter out the cyclical factor.

      Delete
    2. Make sure to keep in mind that the triangle on the balance sheet is presented in calendar year format, rather than accident year format! Because of that, you need to be careful when progressing past the first year in the triangle because it's cumulative. It includes all years prior to that year, as well, so it's not like you could just add up all of the redundancies and deficiencies or something.

      To get a better idea on how they've done in each year, you can convert the calendar year triangle in to an accident year triangle pretty simply, and that'll give you a better idea of how the company has done over time and trends in developments than the calendar year triangle will. :)))

      In regards to paid claims, you can get a rough estimate of the company's reserving pretty simply. Make sure to convert the triangle to accident year, and then check reserving with a simple chain ladder method. You can calculate the company's link ratios over time to get what the ultimate loss development factor is going to be. ^.^

      It's not like it'll be perfect or anything, but it's a cool way to check if reserving is even close to sensible! :D Here's a brief overview of how to do it yourself, if you want. Enjoy~~

      http://www.aicp.net/chapters/midwest/prop_casnotes0922804.pdf

      Delete
  8. I get that, it could just be a bad year for the industry but just the level of remaining reserves seems extremely low given historical developments for those years ('03 and '04). I guess you have to trust that Markel has done their homework. I love the blog by the ways its one of the best I've ever read.

    ReplyDelete
  9. Its amazing how well tobbaco stocks have performed over the years considering all the lawsuits thrown at them. This is a unbelievable success story in many ways.

    ReplyDelete
  10. Is anyone surprised that ALTE is now above 30 dollars a share as of 1/24/2013?

    ReplyDelete
  11. I love the blog by the ways its one of the best I've ever read.

    ReplyDelete
  12. Hi, I can not seem to work out who runs this blog, however someone has deiced to write about my company from jononphp6 which they linked to my website. If the owner could remove this asap that would be very much appreciated.

    ReplyDelete
    Replies
    1. Hi,

      I deleted that post. Please let me know if it pops up again somewhere... I try to delete spam but don't get them all.

      Sorry for the inconvenience.

      Delete