On Price, Intrinsic Value, MBS, and Mark-to-Market

[davelj]

Since I’m watching football and have a little time I thought I’d offer some perspective on a couple of issues that have come up in other threads recently.

One issue is Price vs. Intrinsic Value. One poster said that, “There is NO SUCH THING as intrinsic value,” and then acknowledged that, in fact, intrinsic value existed a concept, but not a useful one because we have prices (which, was suggested, are all that matters). Fair enough. In general I disagree completely, but specifically will acknowledge that the idea of intrinsic value has been bastardized by too many financial institutions via mark-to-model accounting. Belief that there is no such thing as intrinsic value is essentially a belief in the strong(est) form of the efficient market theory.

The concept of intrinsic value is that there is a value to an asset based on its expected cash flows. Then there is a value that the market places on an asset – it’s price – that reflects not only expected cash flows, but also a “speculative component,” comprising expectations about where the security might trade in the short term (according to Keynes’ “animal spirits”) and the asset’s liquidity. The problem with intrinsic value is that it’s theoretical and only known with precision with hindsight; the problem with price is that it’s set at the whim of Mr. Market, who’s notoriously moody and generally wrong by varying degree with hindsight.

Two simple examples:

(1) Let’s say you own an apartment building with 20 units that you believe you could sell for $3 million if you had six months to market it properly. And let’s further assume that your assumptions are reasonable. So, for argument’s sake, let’s assume that the value is $3 million. Now, what would the price be for the same apartment complex if you had a week to sell it? Probably not $3 million. Maybe $2 million. Maybe a little more. But there’s going to be a huge discount placed on that value if you must liquidate it quickly.

(2) The standard deviation for returns on the US stock market has been about 19% annually since 1870. The standard deviation for earnings growth has been about 12% and for dividend growth around 6%. The fact that price volatility for US stocks is 150% of earnings volatility and 300% of dividend volatility suggests that maybe, just maybe, there’s a gap between price and value at work here. That is, the prices of US stocks have historically been considerably more volatile than there underlying values.

I would argue that almost every financial asset is inefficiently priced at any given point in time, the only issue is to what degree it’s mispriced. Of course, we only discover in that in hindsight.

Now let’s move on to MBS, which apparently many here believe to worthless as an asset class (no one offers up any analysis, but there’s been plenty of opinion). I believe that most of these securities were overvalued as recently as 6 months back – many WAY overvalued (and plenty are worthless). So, like most of you, I was offended by the mark-to-model accounting employed by many of the big banks that allowed them to keep similar private label mortgage securities on their books at levels above the market’s assessment. Obviously, times have changed in recent months. So, for a little perspective, let’s analyze one security that I’ve discussed with friends to see what’s imputed in its price.

CWABS Asset-Backed Certificates Trust 2007-7 is comprised of almost 5,000 subprime mortgages with aggregate principal of just over $1 billion. There are two pools, both conforming and non-conforming, and 64% of the Trust is comprised of 1st liens, with the remainder second liens. There are 16 tranches, or certificates, the top 6 of which are rated AAA. In this particular deal, if the subordinated tranches are wiped out, principal payments are apportioned to the senior-most tranches until they are paid down completely. The current quote on the AAA certificates is 56. Three months ago the quotes were in the 80s. The expected yield is about 23% with an expected weighted life of 2.5 years. Now, here’s the critical part: Holders of the AAA certs will achieve this yield if 83% of the Trust’s remaining principal balances are liquidated at 71% loss severity. If the loss severity goes to 85%, the yield drops to the low double digits on the AAAs – not too bad. Now, I’ve omitted a lot of info for brevity, but you can delve into some of these securities yourself here: http://consusgroup.com/v2/landing_pages/companies_sub/C14.asp.

Now, let’s assume for the sake of the argument that these AAAs are mispriced (that is, too low). The question is: Why? The answer is simple: Liquidity. Buying these securities to capture the cash flow yield requires time and, thus, patient capital. A bank that would otherwise buy this security can’t buy it because its price might decline thus impairing the bank’s capital via mark-to-market conventions. A hedge fund can’t buy it because its price might decline, thus impairing the fund’s performance, thus leading to more redemptions and the need to possibly sell into an increasingly illiquid market. For most potential buyers of these AAAs, even if they’re right about the value, they risk getting burned in the short term if they buy these things. The only folks who can buy these securities are private equity funds and other folks with long-term, locked-up capital. And these groups are few and far between and have their own fish to fry. In this environment, illiquidity trades at an enormous discount. Many professionals I know and trust are as highly confident as they’ve ever been that some of these MBS represent the best risk-to-reward they’ve ever seen, but outside of their personal accounts they don’t have the proper (long-term) vehicle to buy and hold the securities. So the selling continues, the prices decline, and there’s not enough long-term capital out there to reverse things. The old negative feedback loop.

I’ve never been a fan of mark-to-model accounting. It’s been a big part of the problem. Having said that, in the current environment, mark-to-market accounting isn’t so hot either. Yes, assets and liabilities that trade in markets with enough liquidity to have meaningful prices should be reflected on companies’ balance sheets at those prices. But the vast majority of a commercial bank’s assets – loans to individuals and business, and other illiquid securities – can’t and shouldn’t be subject to the instant pricing whims of a speculative market. (See Example 1 above to see where that leads.) There’s gotta be some middle ground. Imagine if you put 35% down on a house in Chula Vista in ’05 and now you’re underwater by 5% on a mark-to-market basis and your bank came to you and forced you to sell the asset within a week to make them whole, current payments be damned. That’s a logical conclusion to mark-to-market accounting in the current environment and I don’t think it makes too much sense. I don’t have a good solution here, I’m just saying it’s not as black and white as some here would like to believe.

Now, feel free to debate these issues, but please, please use data and math in making your points. As the good Professor says, “In God we trust. Everyone else bring data.” I have a tough time engaging in how people “feel” or “think” (with no supporting data) about issues, which is what I see a lot of here. I’m here at the Pigg in fits and starts, so I’ll try to participate in this thread in a semi-timely manner but make no guarantees.

[HereWeGo]

Here’s the prospectus.

[HereWeGo]

Here’s the prospectus.

[HereWeGo]

Here’s the prospectus.

[HereWeGo]

Here’s the prospectus.

[HereWeGo]

Here’s the prospectus.

[barnaby33]

I still feel you need a hug. Intrinsic, belonging to a thing by its very nature. If housing by its very nature has an future expected cash flow, then why is housing in Detroit, or portions of it, basically free? My answer is, demand for housing is not intrinsic and therefor neither is its price. Its based on all sorts of factors that change over time, sometimes drastically.

You’ve now made a straw man argument against me. You’ve changed what intrinsic means, but I’m ok with that. What I’m not ok with is your supposition that I must believe in efficient market theory because I don’t believe in intrinsic value.

Let me state up front what I believe and then you can attack it or let it stand as you please. I believe that at an given time any investment(s) value is the sum of all information available to the buyer about that investment, coupled to the alternatives that buyer has either at the moment or in the expected future. That doesn’t mean that there aren’t huge asymmetries in information. It also assumes that there is no past and of course no future to model. That is as close to intrinsic value as I come. The reason I say that its not intrinsic is that past performance is itself based on a series of factors which are constantly changing.

Now to an example of MBS “data.” As you can see this is a 2007 vintage MBS and as of Mish’s last post on it was already 12% delinquent. Why would anyone take a risk on an investment so toxic? Sure there may be real underlying value that can be realized by buying and holding till the market recovers, but the trick is, you have to be understand whats in the pool. Part of the panic out of these instruments is that they were designed to be opaque and facilitate fraud. So far as I know, nobody really understands most of them. If they do, they stand to make a lot of money buying them on the cheap. All accounting systems have their flaws, mark to model as well as mark to market. Defending mark to model though is irresponsible, seeing as how the banks have used it to try and hide not only their losses, but their culpability in creating these monstrous securities. You argue bring data, but thats been impossible. Trust has been destroyed exactly because of the opacity of many of these instruments and the fraud they enabled.

When we can bring data, then the market can clear!

[barnaby33]

I still feel you need a hug. Intrinsic, belonging to a thing by its very nature. If housing by its very nature has an future expected cash flow, then why is housing in Detroit, or portions of it, basically free? My answer is, demand for housing is not intrinsic and therefor neither is its price. Its based on all sorts of factors that change over time, sometimes drastically.

You’ve now made a straw man argument against me. You’ve changed what intrinsic means, but I’m ok with that. What I’m not ok with is your supposition that I must believe in efficient market theory because I don’t believe in intrinsic value.

Let me state up front what I believe and then you can attack it or let it stand as you please. I believe that at an given time any investment(s) value is the sum of all information available to the buyer about that investment, coupled to the alternatives that buyer has either at the moment or in the expected future. That doesn’t mean that there aren’t huge asymmetries in information. It also assumes that there is no past and of course no future to model. That is as close to intrinsic value as I come. The reason I say that its not intrinsic is that past performance is itself based on a series of factors which are constantly changing.

Now to an example of MBS “data.” As you can see this is a 2007 vintage MBS and as of Mish’s last post on it was already 12% delinquent. Why would anyone take a risk on an investment so toxic? Sure there may be real underlying value that can be realized by buying and holding till the market recovers, but the trick is, you have to be understand whats in the pool. Part of the panic out of these instruments is that they were designed to be opaque and facilitate fraud. So far as I know, nobody really understands most of them. If they do, they stand to make a lot of money buying them on the cheap. All accounting systems have their flaws, mark to model as well as mark to market. Defending mark to model though is irresponsible, seeing as how the banks have used it to try and hide not only their losses, but their culpability in creating these monstrous securities. You argue bring data, but thats been impossible. Trust has been destroyed exactly because of the opacity of many of these instruments and the fraud they enabled.

When we can bring data, then the market can clear!

[barnaby33]

I still feel you need a hug. Intrinsic, belonging to a thing by its very nature. If housing by its very nature has an future expected cash flow, then why is housing in Detroit, or portions of it, basically free? My answer is, demand for housing is not intrinsic and therefor neither is its price. Its based on all sorts of factors that change over time, sometimes drastically.

You’ve now made a straw man argument against me. You’ve changed what intrinsic means, but I’m ok with that. What I’m not ok with is your supposition that I must believe in efficient market theory because I don’t believe in intrinsic value.

Let me state up front what I believe and then you can attack it or let it stand as you please. I believe that at an given time any investment(s) value is the sum of all information available to the buyer about that investment, coupled to the alternatives that buyer has either at the moment or in the expected future. That doesn’t mean that there aren’t huge asymmetries in information. It also assumes that there is no past and of course no future to model. That is as close to intrinsic value as I come. The reason I say that its not intrinsic is that past performance is itself based on a series of factors which are constantly changing.

Now to an example of MBS “data.” As you can see this is a 2007 vintage MBS and as of Mish’s last post on it was already 12% delinquent. Why would anyone take a risk on an investment so toxic? Sure there may be real underlying value that can be realized by buying and holding till the market recovers, but the trick is, you have to be understand whats in the pool. Part of the panic out of these instruments is that they were designed to be opaque and facilitate fraud. So far as I know, nobody really understands most of them. If they do, they stand to make a lot of money buying them on the cheap. All accounting systems have their flaws, mark to model as well as mark to market. Defending mark to model though is irresponsible, seeing as how the banks have used it to try and hide not only their losses, but their culpability in creating these monstrous securities. You argue bring data, but thats been impossible. Trust has been destroyed exactly because of the opacity of many of these instruments and the fraud they enabled.

When we can bring data, then the market can clear!

[barnaby33]

I still feel you need a hug. Intrinsic, belonging to a thing by its very nature. If housing by its very nature has an future expected cash flow, then why is housing in Detroit, or portions of it, basically free? My answer is, demand for housing is not intrinsic and therefor neither is its price. Its based on all sorts of factors that change over time, sometimes drastically.

You’ve now made a straw man argument against me. You’ve changed what intrinsic means, but I’m ok with that. What I’m not ok with is your supposition that I must believe in efficient market theory because I don’t believe in intrinsic value.

Let me state up front what I believe and then you can attack it or let it stand as you please. I believe that at an given time any investment(s) value is the sum of all information available to the buyer about that investment, coupled to the alternatives that buyer has either at the moment or in the expected future. That doesn’t mean that there aren’t huge asymmetries in information. It also assumes that there is no past and of course no future to model. That is as close to intrinsic value as I come. The reason I say that its not intrinsic is that past performance is itself based on a series of factors which are constantly changing.

Now to an example of MBS “data.” As you can see this is a 2007 vintage MBS and as of Mish’s last post on it was already 12% delinquent. Why would anyone take a risk on an investment so toxic? Sure there may be real underlying value that can be realized by buying and holding till the market recovers, but the trick is, you have to be understand whats in the pool. Part of the panic out of these instruments is that they were designed to be opaque and facilitate fraud. So far as I know, nobody really understands most of them. If they do, they stand to make a lot of money buying them on the cheap. All accounting systems have their flaws, mark to model as well as mark to market. Defending mark to model though is irresponsible, seeing as how the banks have used it to try and hide not only their losses, but their culpability in creating these monstrous securities. You argue bring data, but thats been impossible. Trust has been destroyed exactly because of the opacity of many of these instruments and the fraud they enabled.

When we can bring data, then the market can clear!

[barnaby33]

I still feel you need a hug. Intrinsic, belonging to a thing by its very nature. If housing by its very nature has an future expected cash flow, then why is housing in Detroit, or portions of it, basically free? My answer is, demand for housing is not intrinsic and therefor neither is its price. Its based on all sorts of factors that change over time, sometimes drastically.

You’ve now made a straw man argument against me. You’ve changed what intrinsic means, but I’m ok with that. What I’m not ok with is your supposition that I must believe in efficient market theory because I don’t believe in intrinsic value.

Let me state up front what I believe and then you can attack it or let it stand as you please. I believe that at an given time any investment(s) value is the sum of all information available to the buyer about that investment, coupled to the alternatives that buyer has either at the moment or in the expected future. That doesn’t mean that there aren’t huge asymmetries in information. It also assumes that there is no past and of course no future to model. That is as close to intrinsic value as I come. The reason I say that its not intrinsic is that past performance is itself based on a series of factors which are constantly changing.

Now to an example of MBS “data.” As you can see this is a 2007 vintage MBS and as of Mish’s last post on it was already 12% delinquent. Why would anyone take a risk on an investment so toxic? Sure there may be real underlying value that can be realized by buying and holding till the market recovers, but the trick is, you have to be understand whats in the pool. Part of the panic out of these instruments is that they were designed to be opaque and facilitate fraud. So far as I know, nobody really understands most of them. If they do, they stand to make a lot of money buying them on the cheap. All accounting systems have their flaws, mark to model as well as mark to market. Defending mark to model though is irresponsible, seeing as how the banks have used it to try and hide not only their losses, but their culpability in creating these monstrous securities. You argue bring data, but thats been impossible. Trust has been destroyed exactly because of the opacity of many of these instruments and the fraud they enabled.

When we can bring data, then the market can clear!

[TheBreeze]

[quote=davelj]
CWABS Asset-Backed Certificates Trust 2007-7 is comprised of almost 5,000 subprime mortgages with aggregate principal of just over $1 billion. There are two pools, both conforming and non-conforming, and 64% of the Trust is comprised of 1st liens, with the remainder second liens. There are 16 tranches, or certificates, the top 6 of which are rated AAA. In this particular deal, if the subordinated tranches are wiped out, principal payments are apportioned to the senior-most tranches until they are paid down completely. The current quote on the AAA certificates is 56. Three months ago the quotes were in the 80s. The expected yield is about 23% with an expected weighted life of 2.5 years. Now, here’s the critical part: Holders of the AAA certs will achieve this yield if 83% of the Trust’s remaining principal balances are liquidated at 71% loss severity. If the loss severity goes to 85%, the yield drops to the low double digits on the AAAs – not too bad. Now, I’ve omitted a lot of info for brevity, but you can delve into some of these securities yourself here: http://consusgroup.com/v2/landing_pages/companies_sub/C14.asp.

[/quote]

Where did you get those loss severity numbers? Since your whole thesis that this TLA is undervalued rests on those numbers surely you have some basis upon which to believe the loss severity is going to be 71% versus 98%? You’ve used math, but as far as I can tell, you’re using “feel” to get the loss severity numbers. I thought we weren’t supposed to do that.

You’ve done a top-down analysis. What I would like to see is someone identify each of those 5,000 properties and then value each of those properties based on some type of equivalent-rent multiplier. Once that is done, you could get a value for the entire pool (minus some percentage for foreclosure costs and what not). Are there people doing this? If not, why not? If there is truly the potential for 25% returns, surely the work involved in dong something like this would be worthwhile, wouldn’t it?

[TheBreeze]

[quote=davelj]
CWABS Asset-Backed Certificates Trust 2007-7 is comprised of almost 5,000 subprime mortgages with aggregate principal of just over $1 billion. There are two pools, both conforming and non-conforming, and 64% of the Trust is comprised of 1st liens, with the remainder second liens. There are 16 tranches, or certificates, the top 6 of which are rated AAA. In this particular deal, if the subordinated tranches are wiped out, principal payments are apportioned to the senior-most tranches until they are paid down completely. The current quote on the AAA certificates is 56. Three months ago the quotes were in the 80s. The expected yield is about 23% with an expected weighted life of 2.5 years. Now, here’s the critical part: Holders of the AAA certs will achieve this yield if 83% of the Trust’s remaining principal balances are liquidated at 71% loss severity. If the loss severity goes to 85%, the yield drops to the low double digits on the AAAs – not too bad. Now, I’ve omitted a lot of info for brevity, but you can delve into some of these securities yourself here: http://consusgroup.com/v2/landing_pages/companies_sub/C14.asp.

[/quote]

Where did you get those loss severity numbers? Since your whole thesis that this TLA is undervalued rests on those numbers surely you have some basis upon which to believe the loss severity is going to be 71% versus 98%? You’ve used math, but as far as I can tell, you’re using “feel” to get the loss severity numbers. I thought we weren’t supposed to do that.

You’ve done a top-down analysis. What I would like to see is someone identify each of those 5,000 properties and then value each of those properties based on some type of equivalent-rent multiplier. Once that is done, you could get a value for the entire pool (minus some percentage for foreclosure costs and what not). Are there people doing this? If not, why not? If there is truly the potential for 25% returns, surely the work involved in dong something like this would be worthwhile, wouldn’t it?

[TheBreeze]

[quote=davelj]
CWABS Asset-Backed Certificates Trust 2007-7 is comprised of almost 5,000 subprime mortgages with aggregate principal of just over $1 billion. There are two pools, both conforming and non-conforming, and 64% of the Trust is comprised of 1st liens, with the remainder second liens. There are 16 tranches, or certificates, the top 6 of which are rated AAA. In this particular deal, if the subordinated tranches are wiped out, principal payments are apportioned to the senior-most tranches until they are paid down completely. The current quote on the AAA certificates is 56. Three months ago the quotes were in the 80s. The expected yield is about 23% with an expected weighted life of 2.5 years. Now, here’s the critical part: Holders of the AAA certs will achieve this yield if 83% of the Trust’s remaining principal balances are liquidated at 71% loss severity. If the loss severity goes to 85%, the yield drops to the low double digits on the AAAs – not too bad. Now, I’ve omitted a lot of info for brevity, but you can delve into some of these securities yourself here: http://consusgroup.com/v2/landing_pages/companies_sub/C14.asp.

[/quote]

Where did you get those loss severity numbers? Since your whole thesis that this TLA is undervalued rests on those numbers surely you have some basis upon which to believe the loss severity is going to be 71% versus 98%? You’ve used math, but as far as I can tell, you’re using “feel” to get the loss severity numbers. I thought we weren’t supposed to do that.

You’ve done a top-down analysis. What I would like to see is someone identify each of those 5,000 properties and then value each of those properties based on some type of equivalent-rent multiplier. Once that is done, you could get a value for the entire pool (minus some percentage for foreclosure costs and what not). Are there people doing this? If not, why not? If there is truly the potential for 25% returns, surely the work involved in dong something like this would be worthwhile, wouldn’t it?

[TheBreeze]

[quote=davelj]
CWABS Asset-Backed Certificates Trust 2007-7 is comprised of almost 5,000 subprime mortgages with aggregate principal of just over $1 billion. There are two pools, both conforming and non-conforming, and 64% of the Trust is comprised of 1st liens, with the remainder second liens. There are 16 tranches, or certificates, the top 6 of which are rated AAA. In this particular deal, if the subordinated tranches are wiped out, principal payments are apportioned to the senior-most tranches until they are paid down completely. The current quote on the AAA certificates is 56. Three months ago the quotes were in the 80s. The expected yield is about 23% with an expected weighted life of 2.5 years. Now, here’s the critical part: Holders of the AAA certs will achieve this yield if 83% of the Trust’s remaining principal balances are liquidated at 71% loss severity. If the loss severity goes to 85%, the yield drops to the low double digits on the AAAs – not too bad. Now, I’ve omitted a lot of info for brevity, but you can delve into some of these securities yourself here: http://consusgroup.com/v2/landing_pages/companies_sub/C14.asp.

[/quote]

Where did you get those loss severity numbers? Since your whole thesis that this TLA is undervalued rests on those numbers surely you have some basis upon which to believe the loss severity is going to be 71% versus 98%? You’ve used math, but as far as I can tell, you’re using “feel” to get the loss severity numbers. I thought we weren’t supposed to do that.

You’ve done a top-down analysis. What I would like to see is someone identify each of those 5,000 properties and then value each of those properties based on some type of equivalent-rent multiplier. Once that is done, you could get a value for the entire pool (minus some percentage for foreclosure costs and what not). Are there people doing this? If not, why not? If there is truly the potential for 25% returns, surely the work involved in dong something like this would be worthwhile, wouldn’t it?

[Author]

Posts