Fairfax

[watsa_is_a_randian_hero]

This is my simple model on FFH.  Does anyone else have anything they do?  I'm assuming we're all talking about base-case values over $400 at this point?

[Uccmal]

Hi Brian, I have normally done something similar.  It was easier to value when NB was still a public company.  With your method I am trying to get my head around a couple of your numbers.

 

Why are you adding 355 M to FFH Asia?

 

Where does the 500 M Runoff come from?  Are you assuming this to be the holdings at Holdco not accounted for by RE, ORH, C&F, and NB?  If so, are these assets not already included in Holdco Marketable securities?  I am not clear if this is a double count?

 

Under the more aggressive secenarios you assign a value to Hamblin Watsa?  How did you get this value?  Is it not double counting if the results of HWICs actions are already in the values of the subs?  HWIC employs a couple of small rooms of employees who have a high value but I am not sure they are worth 800 M?

 

I am hoping I am not nitpicking.  As I said I would value it in a similar way.  I would actually apply 1.3 x the value of each long term subs book since this is what they bought NB in at.  The results would be similar to your base case.  We benefit from the extreme amounts of details Prem provides in the reports. 

 

Thanks, A. 

[watsa_is_a_randian_hero]

Here are some more details on my assumptions:

 

FFH asia the $355 represents the difference between what they've booked for ICICI and what the appraised fair value was. 

 

The runoff is a tough one.  A couple years ago my conservative value would have just book $0 for runoff as a conservative number.  However, recent investment gains have increased the book value of the runoff group so much that I don't think that a $0 value would be overly conservative now.  The $500 represents slightly less than 50% of book.  Part of its book is counted in the equity of other subs.  It also represents valuing runoff at only the value of recent investment gains (saying the conservative value would be $0 if it weren't for recent gains, which will probably be distributed back to fairfax holdco anyways).

 

For Hablin Watsa, I don't think this is double counting.  The value of the subs are based off of 1x, 1.1x, and 1.2x current BV or 1.1x, 1.2x, and 1.3x 12/31 BV (if it is based off of 12/31 numbers, I used higher book multipleto account for recent gains).  This I think would be a normalized market insurance-company multiple.  However, what sets FFH apart is its investment management.  Essentially I've valued FFH as a generic insurance company with generic multiples (assuming a generic insurer cannot generate alpha), then added additional value for HWIC. 

 

The value of HWIC is based off the presumption of 5% growth in investments, 20% discount rate (net 15% cap rate), and $18 billion current investments.  Under conservative I assign 0 value.  Under base case I assume 1% alpha is created above and beyond what a generic insurer would earn.  Under aggressive I assume 2% alpha is created above and beyond what a generic insurer would earn.

 

In response to the question whether a couple of small rooms of employees are really worth $800mm or more, I would pose the question, how much value do you think Prem Watsa alone creates for FFH?  I would argue at least $500 million on ability to create alpha.  I changed to include all HWIC after recent performance on CDS generated by other key members.

[oldye]

"In response to the question whether a couple of small rooms of employees are really worth $800mm or more, I would pose the question, how much value do you think Prem Watsa alone creates for FFH?  I would argue at least $500 million on ability to create alpha.  I changed to include all HWIC after recent performance on CDS generated by other key members."

 

About 5 billion, without these guys the company isn't worth more than book. 

[watsa_is_a_randian_hero]

"In response to the question whether a couple of small rooms of employees are really worth $800mm or more, I would pose the question, how much value do you think Prem Watsa alone creates for FFH?  I would argue at least $500 million on ability to create alpha.  I changed to include all HWIC after recent performance on CDS generated by other key members."

 

About 5 billion, without these guys the company isn't worth more than book. 

 

Well I don't know if I'd go so far as to say $5 billion (you need to account for the risk that these guys are human, and human life in general is fragile).  That being said I would say base-case we should have another 15 good years based on current age of management.

 

Brian

[oldye]

We all have about a .8-1% chance of dying in any given year but not sure how that effects their value.

 

 

Consider that they're running 18 billion and that an extra 5% return on those assets is 900 million a year or roughly the value added.  5 billion is a bargain price! 

[triedtestedand]

Hey ... do you think the employees at HWIC read this board and point to it in requesting annual bonuses?  Seriously though, I would look to value them on a differential basis compared to some benchmark peer group vs an absolute basis ...

[watsa_is_a_randian_hero]

Hey ... do you think the employees at HWIC read this board and point to it in requesting annual bonuses?   Seriously though, I would look to value them on a differential basis compared to some benchmark peer group vs an absolute basis ...

 

I am comparing them on a relative basis, not an absolute basis.  The benchmark is other insurance companies.  In my base-case, I've essentially said they will generate 1% alpha relative to what the average portfolio manger for a P&C company can do.

[oldye]

"The benchmark is other insurance companies.  In my base-case, I've essentially said they will generate 1% alpha relative to what the average portfolio manger for a P&C company can do."

 

Unless the benchmark is Markel and Berkshire, I don't know how you get 1% alpha...

 

The average insurer has grown its capital by 8% over the last couple of decades according to Buffett, Fairfax had averaged more than 23%.  The average insurance company can't invest in equities and hope to earn 20%, or nearly 10% from bonds

[watsa_is_a_randian_hero]

"The benchmark is other insurance companies.  In my base-case, I've essentially said they will generate 1% alpha relative to what the average portfolio manger for a P&C company can do."

 

Unless the benchmark is Markel and Berkshire, I don't know how you get 1% alpha...

 

The average insurer has grown its capital by 8% over the last couple of decades according to Buffett, Fairfax had averaged more than 23%.  The average insurance company can't invest in equities and hope to earn 20%, or nearly 10% from bonds

 

With respect to the statement that FFH has grown capital at 23%, that is not what we are talking about here.  That number is the book equity, which is levered through debt and insurance float.  Actual (unlevered) investment returns are lower. 

 

These are meant to be reasonable projections.  Their 15 year average is 4% on bonds and 10% on stocks per 2008 annual report.  This averages out to 5-6% alpha on investments, depending on portfolio allocation.  However, after subtracting for below-average underwriting (losses of 2.8%, while comparable companies might average loss of 1%), I would say historical performance generated 3-4% alpha, counting both assets and liabilities.  That might not be possible in future given scale of operations today (and I can guarantee the market will never price in an assumption that prem will earn 4% alpha in to eternity); I have used 1%-2% as reasonable projections that the marketplace could use as assumptions to estimate fair value. 

[Kiltacular]

I think the numbers and analysis are solid -- especially as you laid out the three scenarios.

 

Thanks for posting the spreadsheet.

[oldye]

"The benchmark is other insurance companies.  In my base-case, I've essentially said they will generate 1% alpha relative to what the average portfolio manger for a P&C company can do."

 

Unless the benchmark is Markel and Berkshire, I don't know how you get 1% alpha...

 

The average insurer has grown its capital by 8% over the last couple of decades according to Buffett, Fairfax had averaged more than 23%.  The average insurance company can't invest in equities and hope to earn 20%, or nearly 10% from bonds

 

With respect to the statement that FFH has grown capital at 23%, that is not what we are talking about here.  That number is the book equity, which is levered through debt and insurance float.  Actual (unlevered) investment returns are lower. 

 

These are meant to be reasonable projections.  Their 15 year average is 4% on bonds and 10% on stocks per 2008 annual report.  This averages out to 5-6% alpha on investments, depending on portfolio allocation.  However, after subtracting for below-average underwriting (losses of 2.8%, while comparable companies might average loss of 1%), I would say historical performance generated 3-4% alpha, counting both assets and liabilities.  That might not be possible in future given scale of operations today (and I can guarantee the market will never price in an assumption that prem will earn 4% alpha in to eternity); I have used 1%-2% as reasonable projections that the marketplace could use as assumptions to estimate fair value. 

 

Thats fine...again I'm not sure how outperforming stocks by 10% and bonds by 5% for the last 15 years= 1% alpha but then again you take into consideration what the market might give them credit for...which is totally outside my circle of reasoning.    I also think they'll be capable of generating cost free float in the future, so my base scenario is more optimistic. 

[Uccmal]

Your numbers appear solid to me now that I see where you got them from, except perhaps for the HWIC estimates.  But you have covered that off by valuing the longterm subs at 1.1 or 1.2 book.  If I applied a value of 1.3 x book (take out value for NB) to each long term sub I would get a similar number and would leave HWIC out.

 

More than one way to skin a cat. 

 

HWIC as an independent franchise would be easy to value.  Comparable numbers could be obtained using Sprott or Gluskin-Scheff both of whom recently went public. 

 

Both are trading at 500 M market cap in a depressed environment.  Assumming HWIC could line up money to manage (which they could) they could probably IPO next year, or last year, at 1 Billion. 

 

If I do that though I would back that value out of the value for NB, C&F, and ORH since they need to be treated as reasonably well run, but not exceptional enterprises.  I think I already said that in reverse. 

 

The HWIC buy-in was done to give the principals their existing shares of FFH.  John Templeton helped them value the company.  You could also take that value and prorate it by 15% per year. 

[watsa_is_a_randian_hero]

Oldye- that is exactly what I'm doing, trying to estimate under a base-case scenario what a reasonable expectation for the market to give them credit for would be.  While all of want the market to give them credit for their historical alpha and forecast that in to the future, I don't think that will happen.  We will be lucky if the market eventually gives them credit at all.  Buy-side and sell-side analysts in the insurance industry tend to focus on underwriting performance rather than investment performance.