ClearFish Research

Jon Corzine just perjured himself. He stated that he had no knowledge of client assets ever having been pledged as collateral. Yet in earlier testimony and through a number of actions it was made clear that the MF Global business model depended on rehypothecation, and the benefits of the UK branch.

rehypothecation explained here.

update: original Reuter's acticle

Appropriate for Today - Guy Fawkes Day

2011-11-04T12:02:00.001-07:00

http://afawkes.wordpress.com/2011/10/09/a-note-from-your-boss/

This too.

Pandora has a Broken Business Model

2011-07-22T08:39:00.001-07:00

Pandora (P) just listed. Nice product for music discovery, where they attempt to put together a custom music station based on a seed song and continuing feedback from the user. They stream the content over the internet to various connected devices.

They have not been profitable, but a quick look at their numbers makes it appear that they are on the cusp of profitability. Unfortunately, there is a problem. Their costs scale with the number of songs they play, but their revenue scales differently, based on their various attempts at advertising. So far it appears that their core costs for revenue are something like 2.1 cents/hour of served music, and have held fairly steady, whereas their revenue/hour of music served has declined from 3.2 cents/hour to 2.7 cents/hour Q/Q, which is quickly approaching their cost level. So, as the hours of music served scales up, their model gets worse and worse.

They know this. That is probably why they limit "free" users (users subject to advertising) to 40 hours/month, or 480 hours/year. They sell an annual service commercial free for $36/year, which together means that their maximum revenue expectation is 7.5 cents/hour, although they have never shown that they can hit that. It's just where they top out.

In their prospectus they spend a lot of time talking about their need for growth of songs served so as to attract advertisers, but at the same time they limit how much they are willing to serve. A bit of a disconnect there. If growth was good, they would not limit it.

Many of their other numbers presented in the prospectus are suspect as well, with some acknowledged as such in some fine print somewhere but others not. For example, they headline 90M accounts in the US. Bald faced misdirection. There are about 330M Americans, so ca. 1 in 4 has a Pandora account? B.S., and they say it's B.S. later. They say they do about 1 minute of audio advertising/hour and give a number for the terrestrial radio industry that is likely much lower than is actually the case. I haven't heard any audio advertising in hours of listening, so they overstate their serve rate and understate the competition's serve rate, making it seem that they are closer together than than they are. Terrestrial radio advertises as much as it does because it has to to remain viable, so misstating that ratio is misleading. At the same time, they talk about their low serve rate as an advantage. Well, yeah, it's nice for users, but not for investors. They don't talk much about their competitor's models. As a quick example, consider Sirius/XM, that has fixed content costs and a subscription model that scales revenue with subscriber numbers. So every new subscriber costs them little and adds a lot of revenue. That's a viable growth model, and is just the opposite of Pandora's.

Unless a miracle happens (advertisers start flocking to them and paying unprecedented rates under their current model, or they change their model somehow) they will have trouble. I think the IPO was a survival bid.

Update: confirmation

Also, none of this says anything about how the stock will do in this most manipulated of all markets of the last 30 years.

Disclosure: no holdings as of posting date/time

Inadvertent Admission that Keynes Multiplier is <<1, etc.

2009-12-03T10:00:00.000-08:00

It has been widely reported, and widely touted by the administration, that the GAO recently determined that the $787B “stimulus” lead to a 1.2-3.5% GDP increase. On a ca. $16T economy, that is ca. $192B - $560B in GDP increase. Since that increase cost $787B, the multiplier for GDP stimulus from this round of government spending is something like 0.24-0.71. That’s right in line with what I said it was, and nowhere near the fantasy disproved number of 1.6 thrown around by the architects of the “stimulus.”

While I’m on ‘inadvertent admissions via numbers’ here’s another one: It has also been widely reported (that phrase means I’m too lazy to look up the link) that the administration has scored 600K-1.5M new jobs saved or created by the “stimulus.” That works out to $1.3M-$524K per job. (Damn, those must be pretty good jobs!) That’s a number the administration seems to be proud enough of as a “stimulus” even though it’s totally cooked and actually much worse than that. But assuming that number is ok, contrast it with the fact that sending a soldier to Afghanistan costs about $500K- $1M/year. That is certainly employment, we get something for the money, and it’s the same price or less per job. You can’t claim one is fiscally irresponsible without making the same claim for the other, but that is exactly what the administration has tried to do (‘oh the high cost of the war’ vs. ‘the “stimulus” has been great!’).

Electric Car Fuel Shock

2009-08-26T09:25:00.001-07:00

Nice article from the WSJ (subscription required) on coming electric cars and claims for huge mpg numbers (since not much gasoline is used). Nissan Leaf (all electric): 367 MPG; Chevrolet Volt: 230 MPG; compared to Toyota Prius: 51 MPG.

Those numbers make the electric vehicles seem like super winners, but they ignore the cost of the electricity. Here’s the money number: Nissan, using EPA guidelines, figures that “…82 kilowatt hours of electricity are the equivalent of one gallon of gasoline…” Using my PG&E bill, the lowest tier of electricity use costs $0.11/KWH, and the highest tier is $0.37/KWH. I almost always max out every month in electricity usage and have at least some electricity usage in the highest tier (the tiers were set in the 1950’s and the electric companies have tried very hard to insure that they have not been changed significantly since). I have a tiny house, but I have a hot tub, and an office, and that is enough to blow us into the top tier of electricity expense. People in McMansions will have it much worse.

So what does that mean? Adding an electric car is an additional electric load, regardless of time of day. At the lowest fantasy rate, I would be paying $9/gallon equivalent of electricity (82KWH x $0.11/KWH). But that is fantasy, since the real rate will be at the top tier that I have to pay every month, which would be $30/gallon equivalent (82KWH x $0.37/KWH). That makes the plug-in electric car seem like a horrible idea.

But, at $30/gallon, and 367MPG, it would cost $8 to go 100 miles (30/367*100). At the fantasy rate, it would be $2.45 (9/367*100). Compare that to an average gasoline car getting about 30MPG, with gasoline at about $3/gallon, which would cost $10 to go 100 miles (3/30*100), and the price seems competitive. But just barely. Maybe.

CDS Bear Raid - it's all a Hoax?

2009-03-26T11:26:00.001-07:00

Excellent analysis by Andy Kessler in the WSJ (with my take added):

Banks had long term holdings (MBS’s, CDO’s) with 10-30 year time horizons, but generally operate on a day to day basis with overnight funding. To count these and their derivatives as part of reserves, regulators demanded that they buy insurance against the derivatives defaulting. That’s credit default swaps (CDS’s) and AIG is the main player that sold them. Blamo! Drive the price of the CDS up, and it’s taken as a proxy for default likelihood, as well as used to value collateralized derivatives through the Gaussian Copula formula. So, drive the CDS price up, and the collateralized derivatives drop, which must be recognized via mark-to-market (i.e. writing off the supposed losses and hence taking immediate hits to the stock price), and so the bank reserves drop, and then they are regulated to have to raise more capital, which gets harder and harder as their reserves drop and their investment grade decreases (by the rating agencies) through a pro-cyclical cycle.

It’s a beautiful thing. Principally because apparently it doesn’t take much money in the CDS market to influence prices (say $25M or so). So for $25M in the CDS market you can make it appear that GE or Citigroup are close to bankrupt.

What a gorgeous bear raid.

That’s part of how a small change in house prices (the objects of collateralization) plummeted the world financial markets. Maybe it was all a big fake out.

Notes on Causes of Crises

2009-03-02T12:16:00.001-08:00

Kindleberger. – credit expansion

Mechanism – collateralization of debt by banks, moved loans from balance sheet to investment roster so it no longer triggers capitalization rules, and thus allows more lending. Result – we went to a 30:1 leveraged economy from a typical 10:1. 10:1 has historically been handled, 30:1 has not. Direct bank lending accounts for 20% of credit market, the other 80% is various forms of collateralization (e.g. the bank isn’t giving you that auto loan). Collateralized debt securities were purchased and copied worldwide. Legitimization of collateralization was triggered by Fannie/Freddie and rating agencies. Collateralization is the worldwide infection vector.

Consequence: more money sloshing around for hot investors (the collateralizers), who largely played the Yen carry trade funneled into the commodities market. (Oil run-up was consequence, not cause. In 80’s majority (ca. 80%) of US GDP was directly tied to oil consuming industries. Today it’s ca. 20-30% (having been replaced by electricity, which is not mostly oil based (it’s coal and nuclear etc.)). Oil run down is not causative of GDP contraction, but rather symptomatic. Also, oil is priced in US dollars which have been weak during the oil price run up. Priced in any other currency, oil didn’t really go up that much.)

Down Side Floor: moving back from 30:1 to 10:1, a 2/3 contraction. 67% drop in everything is probably the floor.

As bad as that is, it can be made worse. Before credit contraction is allowed to take it’s natural course, we will see (and have seen) policies leading to not-quite-hyper inflation due to a lot of irrelevant drink-the-punch-before-the-bowl-is-empty crazy socialism spending, and just-for-philosophical-fun killing off of a few perfectly healthy additional industries (e.g. healthcare).

Update 12 Mar 2009: Beautiful explanation in Wired of use of Gaussian Copula formula as infection vector for global financial meltdown (the problem with the methodology used for collateralization).

Multipliers

2009-02-12T11:47:00.001-08:00

I’ve taken a long hiatus from blogging, and expect that to continue, but thought I would mention this:

I am just a simple cave man, and perhaps I do not understand your ways, but I do know that whenever you do any sort of modeling (whether it be femtosecond chemical dynamics, protein folding, regional ozone formation, or financial markets – all of which I have done) your model must survive the simplest boundary cases to be prima facie correct. Or rather, if it doesn’t satisfy the simplest boundary cases, then you know it has to be wrong.

The theoretical premise of the current “stimulus” pork package is the “multiplier.” It is estimated by the Keynesian spenders behind the pork package to be something like 1.4 – meaning for every dollar spent by the government, we all get 1.40 dollars back. That is prima facie absurd. If it were so my best investment would be to give as much money to the government as I could, so they could redistribute it, resulting in my getting $1.40 back for every $1 I gave them. I would overpay my taxes to the 100% level because of the easy excess returns. I think you know that doesn’t work. By the way, it was tested – a little thing called Communism, and it sort of failed as an economic exercise. If I am taxed at 100%, I do not expect a 40% return. The multiplier on taxation, i.e. government spending, is more like zero.

At the other extreme, if I am not taxed at all, the one thing I know is that I will keep that dollar. So the multiplier on keeping my dollar (i.e. lower taxes) is something like 1. There are definitely services provided by the government that are worth paying for, so the multiplier isn’t exactly 1, but it is around there.

Any model that is not complete garbage has to have the multiplier for 100% taxation at about 0 and for 0% taxation at about 1.

But I am just a simple caveman.

Update 12 Mar 2009: Academically rigorous evidence (pdf) that the 1.5 multiplier is a fantasy of "old Keynesian" fallacies, when in fact the multiplier is much less than 1. Also, Numbers in WSJ give multiplier estimate of 0.1. ($14.3T GDP * 0.6% GDP growth = $87B growth, vrs. $800B spent -> 0.1 multiplier)

General Market Thoughts

2006-07-07T10:22:00.000-07:00

Is it just me, or is the market a bit shiftless right now?

The IPO market has taken a breather, it has become more difficult for private capital to cash out through the public markets, the technology market has been largely boring (no replacement cycle, hot new product trends, etc.), the ethanol market has taken a dive, the housing market remains ambiguous (consumption high, but producers continue to warn), and there are no large discernible secular trends. Does that mean we are stuck in a sideways market, on the first leg of a major downturn, or basing for a resurgent upturn?

It's hard to say with any confidence, but our sense is that the market is basing. The sell-in-May crowd clearly took over in, well, May. That is typical of most years, but has been a bit atypical for recent years. Interest rates, as well as geopolitics loom large in the market (the latter principally through oil and it's feedbacks). Can either of those get much worse (i.e. interest rates go up much higher, or oil prices continue to rise)? They could, but our guess is that there is not a lot of steam left in either of them.

At some point in the not too distant future, the Fed will pause. We think that will happen coincident with the typical timing for buyers coming back to the market after Summer. We don't expect much excitement until then. Individual plays continue to tick around, but big money flows are probably on the sidelines for another month or two. That means momentum investing is probably moribund for a bit, but it could be a good time to pick up quality, or to short junk.

A Bit of Geopolitics - North Korean Positive Developments

2006-07-05T09:11:00.000-07:00

Just as the ascent of Hamas in the Palestinian Territories provided a clarifying moment to all but the most ideologically locked regarding the Palestinian/Israeli conflict, the test firing by North Korea of a number of short, medium, and a long range ballistic missile will provide a similar clarifying moment. Good faith opposition to good ideas is generally founded on differing interpretations of murky facts, and clarifying moments help to tip the interpretation of those facts toward consensus. The short term may always have serious ups and downs, but we view the North Korean developments as a long term plus. North Korea is only a surviving Stalinist rogue belligerent because of enablers - principally China. Having North Korea as a belligerent served China's diplomatic purposes. That position will now seem considerably less tolerable. Similarly, it is now clear to Japan that their role in the world will be more directly active.

Long term, this means that the North Korea situation has begun the process of rectification, as the world players will begin to agree upon some commonalities, or look increasingly like nutters themselves. And that too is clarifying.

With clarity, eventually comes resolution. You can't solve problems until you agree on the facts. Solving problems may still be an arduous long road, but recognizing the interesting facts is always the first step.

1 Year Anniversary

2006-06-28T07:51:00.000-07:00

Happy Birthday to the ClearFish Research Blog!

If anyone want's to say thanks, please do leave a comment.

Public Storage (PSA)

2006-06-27T13:44:00.000-07:00

Story: A long-term reader asked about Public Storage, citing in particular that it's prospects may be a proxy for overall GDP growth, and adding that as people get older they acquire more stuff which they find difficult to get rid of. That highlights a possible demographics effect for PSA's prospects. The stock has been on a 3-year steady uptrend, with a recent ca. 12% pull-back. I have always found investing based on broad demographic changes to be difficult, as shorter term exogenous factors, as well as the usual business management factors, can derail the effectiveness of the overall shift. We'll just acknowledge that demographics probably are in PSA's favor, and see if we can make a case for them based on shorter term criteria as well.

There are a couple of things we would anecdotally add to the story, before launching in to the company: First, it is easier today to efficiently get rid of stuff via eBay or Craigslist, but that isn't as true of large items like furniture. But the disincentive to divest isn't always motivated by economics, as opposed to psychology. Storing, as opposed to selling, allows for putting off those psychologically jarring decisions to a later date. Second, there has been some anecdotal discussion of a shift from "large homes" (4000 sq. ft.) to "smaller homes" (2500 sq. ft.). By historic standards, both of those are huge, and both provide ample storage space. But a cluttered 4000 ft. home will not fit in a 2500 sq. ft. space, so retiring or downsizing baby-boomers may help the storage market. On the other hand, a lot of that excess stuff can go to the kids as they get set up, or moved to the vacation home. Both of those trends are likely slow moving ones.

Company: From their latest quarterly report (pdf) ending 31 March 2006, they have had Y/Y quarterly top line growth of 11% ($278.7M/$250.9M), or $28M. Of that, $2.2M, or ca. 8% ($5.075M-$2.893M) could be directly attributed to increased interest rates. That says 90% or so of their business is not interest rate dependant in a direct fashion. However, to the extent that a rise in interest rates slows the upscaling of homes, an interest rate increase may indirectly be good for their business. Similarly, to the extent that an interest rate rise stimulates downsizing of home, it would also work to indirectly augment their business.

Y/Y quarterly expenses increased 7% ($161.4M/$150.7M), or $11M, showing that revenue is growing a wee bit faster than expenses, for an overall Y/Y quarterly net income growth of 18% ($114.2M/$96.4M), or $17.8M, and a quarterly EPS of $0.48.

That all looks fine, but it's a bit boring.

If we look at their latest annual report (for 2005, pdf), and compare their numbers for year end 2003 - just prior to the current run-up in the stock - to the year end 2005 numbers we see that revenue has increased 19% ($1,061M/$894M), expenses have increased 7% ($604M/$565M), and net income has increased 35% ($456M/$337M), or absolute increases of $167M, $39M, and $119M, respectively (they have a few other line-items that account for those number not adding up). EPS have increased 54% ($1.97/$1.28).

Stock: At a 2003 year-end price of about $30/share, they had a trailing P/E of ca. 23 ($30/$1.28). For year end 2005, at a price of about $70, they have a trailing P/E of ca. 36 ($70/$1.97). A 2003 P/E applied to their 2005 earnings would give them a share price of $45 ($1.97x23). So, it seems that most of their share price increase has been due to a market multiple expansion, rather than fundamentals. That is perfectly reasonable if justified by their growth prospects, but we think they have gotten ahead of themselves. With Y/Y earnings growth in the 20% range or so, a market multiple of 23 seems about right. The current value of 36, we're guessing, is due to the demographic story without running the numbers. They are a Solid business, but we think their stock price deserves a pullback.

No Matter What You Do, You're Wrong

2006-06-26T09:31:00.000-07:00

If:

You buy on the way down, and it keeps going down (should have waited), or
You sell on the way up and it keeps going up (should have waited), or
You buy on the way up and it keeps going up (you should have bought earlier), or
You sell on the way down to stop losses and it turns around, or
You don't buy at all (opportunity loss), or
You don't sell at all and hold those losses, or...

You can always view whatever action you take (profitable or not) as poor.

The perfect is the enemy of the good. In this business no matter what you do you can beat yourself up over it. Only if you call the bottom and sell at the top can you really feel good about it, but even then you could be wrong on a different time frame.

That investor psychology issue (as well as profits) plays a big role in the attractiveness of momentum investing - buying on the way up and selling somewhere higher.

Not actionable information, but nevertheless it's good to remember that if one is so inclined, one must somehow escape any inherent negativity to survive in this field, as one can't optimize perfectly. Go for profits without perfection.

Housing Market

2006-06-26T08:56:00.000-07:00

The US Commerce Department released their report on housing sales (pdf) for May 2006 today. The headline in the WSJ is that they are unexpectedly up 4.6%, whereas "economists" were predicting a 4.0% decline. 1) Economic predictions of housing gyrations have been poor, and they continue to be so. 2) The numbers behind the headlines show something a bit different.

The seasonally adjusted numbers are what always get reported. In the latest report they give actual sales for 2005, as well as annualized numbers from the monthly snapshot for 8 different months. The annualization process has a lot of slop in it, and I have previously urged readers to discount it and just look at the actual numbers. Another way to use the annualized numbers is to put error bars on them. The actual number of homes sold in 2005 was 1.283M. The monthly annualized numbers from May through December vary from 1.236M to 1.367M. That says that the error bar on an annualized number should be at least 0.09M, or 7%.

That says that any variation in housing numbers that is less than 7% is just noise in the annualizing algorithm, and signifies nothing.

Similarly, if one looks at the error bars in the regional numbers, the variations due to the annualizing algorithm are always less than the error bars. All the sturm und drang over the housing market is noise in the annualizing algorithm.

For 2006, the latest annualized is 1.234M houses sold, which really means something between 1.148M and 1.320M, so 2006 home sales could be less than or greater than 2004 and 2005 homes sales.

Without seasonal adjustments, 2006 year-to-date home sales are down 11% from 2005 sales, but that means almost nothing as the summer sales season is not incorporated in that data yet, and month-to-month sales vary significantly due to exogenous (weather, gas prices, world events, etc.) and random factors.

The housing market implosion is overblown, or even non-existent.

Google's new Pay-for-Performance program

2006-06-22T10:10:00.000-07:00

It's always nice to be right, even if you don't fully expect to be. I harped for a while (here for example, and here too) about the disadvantages of Google's pay-per-click advertising model (advertising dollars change hands every time an ad is clicked), arguing that the trend for the future would be a transition to pay-for-performance (advertising dollars change hands only after an ad-stimulated purchase occurs). Now the WSJ (via the AP, via Seeking Alpha - of which I'm a contributing member) reports that Google has just such a program in place, deemed "cost-per-action." Kudos to Google for getting ahead of the transition and setting themselves up to own the space.

Such a transition has serious repercussions for the online advertising marketplace - for advertisers (more efficient use of dollars, better marketing budget cost controls, no click-fraud), ad-supported venues (less widely distributed ad revenue), and competitors. Google will now be a three-tiered advertising player, representing the full spectrum of online advertising evolution from pay-for-impression (CPM), to pay-per-click (CPC), and finally to pay-for-performance (CPA).

I thought the transition would take a couple of years, but I guess in internet time a few months is a couple of years.

[All previous Google advertising comments]