Monthly Archives: May 2012

Thoughts on portfolio allocation

With the recent addition of Origen Financial the number of companies in my public portfolio reached the magical number ten, and I thought It would be a good idea to talk a bit about my ideas on portfolio allocation.


The essence of my thinking is simple:

Don’t put all your eggs in one basket.

While I guess most readers would agree with this statement, it probably also means something different for everybody reading it. Some people think you should concentrate your money in your best idea’s and owning a half dozen companies is enough to not have all your eggs in one basket. Others might want ten times as many baskets.

I think both approaches can both be right and wrong. While it’s not irrelevant the number of companies you have doesn’t necessarily say how diversified you are. You can own a lot of different companies, but if they are all located in the same country your geographic exposure might be very concentrated. But just looking at where the company has headquarters is not enough: there is a big difference between a company that only serves the local market or one that has a global reach. If you buy an oil major it’s nearly irrelevant in what country it’s headquartered.

It also matters how big and risky the company is: a company that has just one niche market, be it local or global, is probably more vulnerable than a company with multiple divisions and countless different products. If market conditions for one product develop unfavorable it doesn’t have to be a huge problem for the big company while it can mean the end for a smaller company.

What this means in practise: I have a fair amount of micro-cap companies in my portfolio, and I don’t think these are suitable for large outsized bets and a super concentrated portfolio. Especially not since there is almost always something ugly in the story of cheap stocks. They might have a strong balance sheet, but they are almost never strong companies with a great competitive advantage. And if the business does deteriorate you never know how much good money management will throw after bad.

Another thing to look out for is that you don’t want too much companies that have their fortunes linked because of some underlying economic force that might not be directly obvious. Two companies can be active in total different industries; but if they are both small and highly leveraged they are probably both getting in trouble in a global liquidity crisis. Or maybe companies are linked because government spending is the biggest source of revenues, or maybe because they are both exporting to the same country.

Especially if you use screeners to find investments you might want to be careful that you don’t end up with a lot of similar companies. Past year value screens used to find a lot of Chinese RTO stocks, and I think you can guess what would have happened if you would have bought a bunch of those stocks.

I don’t think diversification means that you need to have exposure to everything. I don’t care if there is for example a specific industries that I have zero exposure to: as long as I have exposure to a sufficient number of different industries I’m happy.

Keeping cash

A lot of value investors believe that you need to keep a lot of cash so you can capitalize when the market offers great companies for bargain prices. I’am not doing this, because while it’s very nice to have cash when the market is dropping it certainly not a free option. If I’m finding companies that I expect will on average generate a >10% return I prefer to put my money in those than in cash that will generate a near zero return. If there is some truly great opportunity I can always sell some of my existing holdings.

And who knows how successful I’m going to be to buy those bargains. Maybe I’m going to buy too early or too late, in both cases reducing the value of the option. I’m not going to try to time the market. I guess I could end up with a lot of cash if there is simply nothing I want to buy, but that’s so far absolutely not the case.

Closing thoughts

There are a lot of subtleties and nuances with regards to portfolio allocation, and there are also some major subjects I haven’t talked about at all. What to think about different asset classes; should I own bonds? Or what about currency hedging? Going to leave that for another post, but hopefully I’ve given some insight in my thinking and strategy.

Some good posts at Greenbackd

The past few weeks has published a series of articles about value investing; why it works or doesn’t work, what kind of statistical measures for cheapness perform best and what not to do if you want to outperform. Since I’m assuming that most readers of this blog are value investors this should be interesting. The articles:

That should be enough reading material for the weekend! And to conclude this post: a nice visualization of volatility for the past two decades based on series of theoretical VIX indices at different maturity levels:

London Finance (LFI.L)

Reading my blog you would probably get the impression that almost every time I research a company I end up buying shares. That’s not really the case, but often I find early on in the research process reasons for not wanting to own the company and I simply move on to the next investment. But today I’m going to do something different and explain why I’m not interested in buying London Finance (LFI.L).

The company was mentioned today at Whopper Investments and it caught my eye because at first sight it is trading at a big discount to NAV. The company has a 5.6M GBP market cap while it has 12M GBP in assets. It has the following assets:

  • £5.4M in shares of three small UK companies
  • £4.2M in a portfolio of UK and European blue chip equities
  • £2.6M in cash

This was enough to get my attention, but there are some good reasons to discount this value. First of all the company does have almost £2M in debt reducing the net asset value. The company reported 34.6p in NAV/share on February 2012 and is now trading at 19.5p, a discount of 44%. European stock markets are down ~10% since, so the current discount to NAV is probably more something like 35%.

The discount is not only not as big as it first looked, more important are the costs of the London Finance entity itself. Last year the company had £749,000 in administration expenses, a huge number compared to the total NAV. They did earn £398,000 in management services fees (not sure where this money is exactly coming from, since as far as I can tell they only manage their own assets?), but even if the company has just net costs of ~250K a year this would translate to a 2.5% expense ratio (and they also have some tax costs as well that are not included in this figure).

You can argue what kind of discount would be fair based on the yearly costs, but if the long term return of the portfolio would be around 10% a fair discount would be around 25%, and I think that a return like that is pretty optimistic these days. Combine that with the fact that the current discount is somewhere around 35% today and I don’t think there is really a sufficient margin of safety available. Especially not since I don’t really understand where the management services fees are coming from. Without those the cost structure would just be beyond terrible.



Origen Financial (ORGN.PK)

Origen Financial manages a portfolio of manufactured housing loans and asset backed securities (ABS). The company delisted in 2008 and is in liquidation mode since. It sold the loan originating and servicing business and is now only managing seven separate trusts that were created between 2004 and 2007 and hold housing loans. The trusts have debt that is non recourse to the parent company. Because some trusts are performing well while others have suffered substantial losses the consolidated balance sheet is useless to evaluate the available asset value, creating a potential opportunity.

The idea shares many similarities with Gramercy Capital that I have written up before, but in the end that one was just too hard to value for me (I’m considering giving that a new shot though). Origen reports monthly performance numbers for the ABS trusts on their website making it easier to value the assets of the company. Since almost all the assets of the company are within those trusts they offer the key to determine the asset value of the company.

Company structure

As already explained Origin Financial holds interests in seven different trusts. The trusts are separate entities that have non-recourse debt. While some trusts are performing and producing cash flow to the parent company other trusts are underwater and are probably never going to produce any cash flow. The cash flow from these trusts is one way traffic: it can only flow from the trust to Origen Financial, not the other way around.

  • 2004-A, 2004-B and 2005-A: these are performing fine and produced 11.7M in cash flow to the parent company the past twelve months.
  • 2005-B: hasn’t produced any cash the past years, but probably does have some value since it is close to meeting requirements.
  • 2006-A, 2007-A, 2007-B: no value?

Each trust started with several layers of debt and a pool of loans with a face value higher than the total amount of debt. To take the 2004-A trust as an example: this one started with $200M in debt that was backed by $239M in face value of contracts with an average contract rate of ~10%. The debt is sliced in several layers. The most senior layer has the lowest interest rate, but it does have the highest rating because when loans default these defaults are first absorbed by the excess amount of collateral and then by the more junior debt levels.

Origen Financial doesn’t own the debt of the trusts (with one small exception) but the remaining equity, so to get cash the interest on the notes needs to get paid first and certain targets need to be met. Most importantly there needs to be an amount of excess collateral available before distributions to Origen Financial can be made. Since all defaults are absorbed by the overcollateral the difference between the interest paid to the debt and the interest received on the total loan pool has to be big enough to absorb these defaults. The equity interest in the trust is leveraged to the default rate of the underlying contracts, and when it’s spikes any excess overcollateral can be wiped out easily. This is exactly what happened with the 2006 and 2007 trusts.

All the trusts start liquidating as soon as they are created. Loans get repaid, default or reach maturity and when this happens the note holders of the trust get repaid as well (starting with the most senior notes). To take the 2004-A trust again as example: it started with $200M in debt and $239M in contract face value, and at the moment $55.7M in debt is remaining and $81.3M of contract face value. When the amount of debt is dropping the amount of overcollateralization that is required is also dropping. The trust started with an overcollateralization target of $42.8M, and the target is now $25.6M. With the target dropping the excess money can be distributed freely to Origen.

So to summarize there are two sources of value:

  • OC targets are lowered, making it possible to return capital to Origen
  • The underlying loans can produce excess cash flow after servicing the debt

Asset value

To determine the value of the interests that Origen has in the trusts the amount of overcollateralization is a dirty shortcut to get some idea of the available value. The 2004-A, 2004-B and 2005-A trust have a total of $62.5M in excess collateral, but this is understating the value of these trusts. For the past three years these trusts all have produced cash flow in excess of the return of overcollateral. This basically means that the current default rate is lower than the delta between the rate paid to the ABS note holders and the rate that the underlying assets are paying. The table below summarizes the performance data from the three performing trusts.

As is visible the amount of required overcollateral dropped by a total of $9.2M the past twelve months while a total of $11.7M was distributed to Origen. So taking the total overcollateralization for these trusts as the total asset value is too pessimistic. At the same time it’s too optimistic for the 2005-B trust. This trust has significant OC ($15.4M), but here defaults have been eating away at the excess collateral for the past years. The graph below shows the OC target, the actual OC and the delta between them.

As is visible the trust is very close to meeting it’s OC target, but so far the available amount of OC is dropping at a nearly identical rate as the target. If this trend continues the trust would have at the end of it’s life no OC available, and it would not have produced any cash flows to Origen. It’s highly likely that this trust does have some value, but it’s almost certainly a lot less than the current amount of OC available. Usually the default risk decreases when the portfolio ages because at the end you keep the people that have been paying you back for years, and at the same time the loan to value ratio is on average going down because principal payments get made.

Answering the question how much value this trust has is very hard. Not only do you need to estimate how much of the overcollateralization is actually excess collateral and will be returned to Origen, the timing of these payments also matter a lot if you want to determine the current value. A scenario where the trust would start to create some cash flow this year would be way more favorable than a scenario where the cash only starts flowing after 5 years. I’m going to take the easy and conservative option and simply ignore any potential value this trust could have.


While we have determined that the value of the assets is significantly above the $62.5M that it has in excess collateral in the three oldest trusts this does not complete the valuation story. Equally important is the timing of the cash flows, and this is the second key part of the thesis. All notes have optional redemption clauses that make it possible for the servicer to buy the contracts from the trust and repay the notes at par when less than 20% of the debt remains outstanding. The graph below shows the amount of debt outstanding for the 2004-A trust, and the optional redemption threshold:

Just by looking at the graph it should be clear that the optional redemption threshold is probably going to be reached somewhere in 2014, maybe 2015. And while it’s called an optional redemption, I think you can be confident that it’s going to be used, since the prospectus states that when the servicer fails to exercise his purchase option an auction process is started to sell the trust assets at the highest possible price. But that’s probably not necessary since the trust holds loans that pay around 10%, so buying the assets and retiring the remaining debt at par should be an attractive transaction in the current low-rate interest environment. The story is roughly the same for the 2004-B en 2005-A trusts: these should reach the thresholds a bit later, probably somewhere in 2015. When this happens the remaining OC is returned to Origen Financial.

Completing the valuation picture

To complete the valuation we have to make some rough estimates about the cash flows and the timing. Since that’s not easy I’m going to take a very simplistic and very pessimistic scenario and show that even in this case the NPV of the assets is higher than the current market capitalization of the company. Assumptions in this case:

  • The three performing trusts are all liquidated after 4 years (begin 2016)
  • No cash is returned before this date
  • The amount of cash returned is the total amount of OC (64M)
  • Discount rate is 10%

If you would throw this in Excel you will get a NPV of $42M while the current market capitalization is $36M. This scenario is especially pessimistic because in reality a lot of cash would be returned earlier, although it’s hard to model this exactly. In 2011 $11.3M in cash flow was generated while $10.9M was distributed to shareholders. You should probably asume that this amount shrinks slowly every year going forward, until the redemption thresholds are reached.

At the same time this valuation ignores the overhead costs of the Origen Financial entity itself. These should be around $2M a year, but the excess interest received above the return of OC should easily be able to cover these costs ($2.5M was received the past year). Another minor source of value is the fact that the company holds $3.5M in notes: probably the B2 notes in the 2005-B trust since these have exactly $3.5M in face value outstanding. These notes have a 7.2% yield and are protected by 15.4M in OC, so these should probably be at least worth par (and should add $0.14/share in value).


The company delisted in 2008 and is only publishing financial statements since, so I don’t know how much insiders exactly own today. In the 2008 annual report insiders and directors as a group owned 37.2% of the company. More interesting is the Succes Fee Letter Agreement with the CEO that gives him incentive to work towards liquidating the company since he will be paid a 1% bonus based on the value of a liquidity transaction.

A little bit of history

Origen Financial got in trouble in 2007 in the height of the financial crisis. It had to sell the loans it owned at large losses (before creating an ABS trust it needs to own a bunch of loans first), it required a rescue loan, and it sold the loan originating and servicing business. At the moment the company doesn’t have any debt at the parent level anymore, and it started distributing liquidation payments in 2010. Origen paid $2.6M in 2010, $10.9M in 2011 and $3.9M has been paid so far in 2012. So this shows that liquidation payments are at the moment occurring at a rapid rate.


The valuation work in this write-up has been crude for assets this complicated, but I think it’s sufficient to show that the asset value behind the Origen Financial is not only significantly more than the current market cap, there is also a strong catalyst in a few years time when the trusts reach the ‘optional’ redemption thresholds. At the same time management has shown that it is willing to liquidate the company and return capital.

While I do think Origen Financial is a compelling opportunity, it’s not without risks, and I wouldn’t want to buy a huge position. There is a lot of leverage at work, although that’s partly offset by the fact that money extracted from the trusts can never flow back.


Author is long ORGN.PK

Asta Funding Q2 2012 results

Asta Funding (ASFI) announced financial results for the second quarter of fiscal 2012 today. The company disappointed in the previous quarters because of the lack of share repurchases after it announced a $20 million discretionary buyback program in the summer of 2011. It switched to a non-discretionary stock repurchase plan on March 9 and it seems that this move has had the intended effect since the company managed to repurchase 116,438 shares at a cost of $923,000 in just three weeks (the quarter ended March 31). Hopefully the pace of the buybacks has remained the same this quarter.

Asta Funding remains very cheap, and hopefully stays that way in the foreseeable future since this would increase the value of the share buybacks for remaining shareholders. The company has $116.4M in cash and securities which is close to it’s current market cap of $132 million. At the same time the company remains profitable, and cash flows from fully amortized portfolios (remember these assets have zero book value!) was $9.25M this quarter, actually up a bit from $9.05M previous year.

One big development of ASFI is the move in personal injury financing. It’s still way too early to tell how this is going to work out, but the company is starting to see the first revenue from this business. This quarter approximately $492,000 of income came from the Pegasus Funding joint venture, and so far the company has invested $8.3M in this business the past six months.


Author is long ASFI