1) Partner Re:

Any way you cut it, Partner Re is Exor's most important business not FCA. If you take Exor's equity method valuation, Partner Re makes up for about a bit over 40% of the entire business' value. I believe that if you foremost care about long term sustainable earnings power, Partner Re is likely closer to 50% of the value of Exor.

What is Partner Re?: We know Partner Re is a re-insurance business and we also know from Buffett that reinsurance businesses due to the low interest rate environment around the world and more competition will not have an easy time over the next decade. But all reinsurer's are not the same, Partner Re looks as follows:

If you dig up Partner Re's financials, you can see that as of March 31^st, 2016, that its balance sheet had USD 15.5 billion in investments and cash, while it had a tad less than USD 1 billion in long term debt. On the underwriting side, unpaid claims are USD 5.3 billion, while long term estimated insurance liabilities are at USD 6.1 billion. I think the first thing you see is that Partner Re is not a reinsurer where you have a ton of liabilities and an equally large amount of assets with a tiny sliver of equity, where if you underestimated your liability side, you have nothing left -> Partner Re has a LOT of equity. 

I think the right way to look at this is to consider all the liabilities that would need to be covered in the short term as something that you cannot earn a return on (you need to have this cash to pay liabilities, you can’t invest in stocks)-> this is the USD 5.3 billion of claim costs and (my very rough estimate) about 1/3 of the estimated insurance liabilities (I assume 3 year average duration), which is about ~USD 2 billion –> total of ~USD 7.5 billion cannot be invested in return seeking assets but must be duration matched to meet underwriting needs. That leaves us about USD 8 billion of Exor balance sheet John Elkann can essentially invest (hopefully better than Partner Re has historically); there is also a small cost to the USD billion debt we have to consider. (all details are in the GeneralRe 10K for 2015).

Investment returns: Partner Re, if you look at the Consolidated Statements of Operations (p.141/1079 in the 10K), has made strict investment income (from bonds I assume) of around $ 450 million/yr in the last 3 years. It has in those years made very little if any (negative in 2 of 3 years) gains on the capital appreciation/loss on invested assets.

If you take the previously mentioned $ 8 billion capital I considered flexibly investable and just say that you can make 10% a year on it, then you come to $ 800 million in investment gains, which is a lot more than the net of their last 3 year’s of investment income + gains/losses from investments. This assumes the other $7.5 billion of assets make nothing.   

Underwriting: If you look at the reserve development tables for PartnerRe (on p. 18 of previously noted 10-K), you can see that they have estimated in most years very conservatively and reserves have developed where it was over-reserved rather than under-reserved. This is good. When looking at combined ratios over the last years; for at least the last 3 years it has varied between 95% and 100% (p. 141 of 10-K) – Note that Partner Re’s reported “Non-Life Combined ratio” were a lot better than this, but Life/Health was negative on the whole. If you assume they can keep an average combined ratio of 96%-98%, you can assume PartnerRe making about another $ 100m - $ 200m on underwriting alone. If they can be more like their Non-Life insurance business/limit the life losses, this number would be a lot better.

The overall point is that Partner Re in all aspects seems to be a conservatively managed and well-run business that can generate significantly more earnings under the right owners than they had in the past few years.

If we take my admittedly very rough estimate above, although I believe it’s rather conservative, Partner Re is a business that can generate ~ $ 1 billion in operating earnings a year.     

Note that I haven’t even considered the value of the part of the float, where the liabilities actually do not need to be “asset matched” and could be invested to get more returns i.e. I think my assumption of $ 7.5 billion that cannot be invested in return seeking assets is conservative. I have assumed the balance sheet holdings easily meet the Solvency II capital requirements under my assumptions, although I have not checked in detail.

2) FCA:

Under my own calculations, the value of this stake makes up roughly 25% of the value of Exor, which is significantly less than the value suggested if you just took the % of equity method valuation by Exor.

Fiat has been written up extensively, so I will just add a few of my views overall.

I see Fiat Crystler, like almost all mass market auto manufacturers as a very capital intensive business, with low ROTCE, and very unpredictable earnings. I think those that want to point the finger at Jeep and say lots of brand value worth X, Y, are trying to find reasons for wanting to say it’s a good business, when it should be seen for what it is: A inherently mediocre quality business with poor economics and lots of debt and pension/union issues. (this is not me saying Jeep is a terrible brand; I think it has potential, but from an investment perspective, this is how I would see it).

The scale it has is certainly less than those of competitors like VW or Toyota (~ 10 million vehicles produced/yr.), or even GM, Ford (~ 5 or 6 million).

FCA produces about 4.5 million vehicles a year, but 2 million thereof are large trucks, which don’t really share scale fully with cars.

Sergio Marchionne is amazing and has done amazing things with this business, and having him certainly is a lot better for this business than not having him at the helm.

His statement that he could take $ 5 bn/year out in costs if he merged with GM or a producer with equal scale is probably right, but such a merger is not easy to find and execute.  

If we assume over a long run this business will make a 2% net margin on revenues (I find this not a very assuming assumption), then we are talking about $ 2 billion in earnings, which values the business at < 5x at today’s valuation. Remember that Exor owns ~30% of this business.

I think telling what FCA’s earnings will be in the next 5 years is anybody’s guess. You have to assume there are tail risks of actual solvency issues possible. But besides that, I find 2 points important to note: 1. It’s clearly not expensive compared to its earnings in several different possible scenarios. And about 1/3 of what Exor holds are actually convertible bonds that are safer and should be seen differently than equity i.e. perhaps like some of the Buffett investments into Salomon or USAir (if I remember correctly).  

3) CNH

CNH is worth 10% -20% of Exor depending on your calculations. It has debt, the agricultural equipment/tractor sector is in a cyclical low point; and CNH is in a 3 way market with Deere and Agco.

I think the economics in this business are much better than the automotive business: there are distribution networks, brands, aftersales etc. involved.

There is quite a bit of debt.  

4) Ferrari

About 10% the value of Exor, I would actually argue that this business in fact does have a lot of brand value. I think Ferrari is genuinely a luxury brand – all economics point it, whether the margins, the limiting of supply, or returns on capital.

Fantastic business, high price (we could argue anywhere from fair to very overvalued, but I think few will argue with me saying this is dirt cheap).

5) Economist/Juventus/etc.

Some of these are nice businesses, but all in all they are too small to really move the needle, so I won’t discuss them.

SUMMARY on the Exor business itself:

I think if you take all the businesses together that comprise the value of Exor, it really should be seen as a lot of cash still to be invested (held currently at Partner Re), and the remainder being an ownership in a collection of mediocre to okay businesses that are currently valued not too expensively.

So, in other words, if you were to compare it to other investment holding companies with managers that follow the Buffett style of investing (John Elkann does), then you have a case where you still have lots of uninvested capital. The capital that has been invested so far is unfortunately not all invested in Buffett quality businesses, but the amount still to be invested can be and I expect it to be more and more moving into this direction in the not too distant future.

I don’t expect Exor to ever completely divest businesses like FCA, CNH, or Ferrari due to history; but the bigger part of Exor even today is not this. And I do fully expect that part to be invested well and into high quality compounders. The reason why I believe this will happen is in the next section.

You know the discount to NAV; the SOTP picture doesn’t look that different from this, although more of the value skew towards the PartnerRe business in my own calculations.

II.             Buying a Partner Re doesn't make you a Berkshire:

While i completely agree with the statement that merely buying a reinsurer doesn't make you a Berkshire, and even note that buying reinsurance business has become in vogue with even many hedgefunds, i believe the crux of this argument is wrong because John Elkann isn’t just anyone, and Partner Re isn’t just any reinsurance business.

John Elkann and the Exor team, I would argue has a solid track record in investments and do understand what quality businesses and compounders are.

If you look at the past businesses they have invested in and have analyzed those businesses yourself, I think the evidence is pretty good:

SGS: SGS is a global testing and inspections (TIC) business in a global oligopoly with Intertek, Eurofins, TuVs, ALS, and a couple other players in different niche subsectors. All in all because of the local nature of regulation that drive a lot of the TIC work, the scale of labs, and relatively low asset intensity, these are high return, fair growth businesses – i.e. compounders. SGS has historically been well managed and the investment Exor made and when they sold was timely.

Cushman & Wakefield: I know this business a bit less well, but in general these types of letting/real estate businesses are fairly decent businesses when run well, and again asset light and good growth – compounding economics. The investment in C&W checks out for me in terms of the typical investment I would expect Exor to make.

There are several other investments, but I think these two, which were fairly significant for Exor pre-Partner Re are reflective of the kinds of investments the company makes. Remember: Fiat, CNH, etc. were not made as investments by the Exor team but were legacy passed down.

Putting the evidence together with what I know about John Elkann/Exor from interviews and accessible documents (annual letters), I basically believe that he is a great capital allocator with a track record that is convincing to me.

The second part of why I think this concern is not correct in the case of Exor is because Partner Re is not just any re-insurer. I think as I have made it clear in the earlier part, Partner Re is very well run, with a history of conservative underwriting. While with lower interest rates, investment income from bonds will can be expected to decline, the fact that the business has so much free capital that could be better invested creates a big opportunity for improvement in earnings.

Couple this with the fact that being a long duration business of Partner Re with a very long historical track record of under-reserving (seen from the reserve development tables), and you can make the argument that reserve releases alone will provide some tailwind to earnings in the next few years.

In any case, I just want to show the contrast of the Exor case to a typical mediocre-run reinsurer with little excess free capital, combined ratios > 100, that have a mediocre investor at the helm: they neither have the possibility or ability to generate good returns from investments or underwriting.

Lastly, I didnt intend to initially write about Brexit, but given its recency, Partner Re is about 40% North American 35% European, Exor is denomiated in EUR.