I’ve never made any secret of the fact that I’m an auto industry analyst only in the most general sense of the term. In other words, I know what I know and I know some people who know what I don’t know. As they know, I need a good three or four renditions of the same financial information before I can even begin to get a grip on its scope, scale and importance. To wit: one of Our Best and Brightest sent me this little tidbit—Chrysler Financial’s Auto Securitization Trust 2009-A—with some explanatory notes. I’m still not sure what to make of it. You? [thanks to you know who you are]
The key takeway is on Page S-10 where they summarize that the raised $1.263 billion uses $1.641 billion of aggregate principal balance of vehicle purchase receivables owed to ChryslerFinancial (77,730 cars).
Page S-20 is the most interesting… basically it breaks down the $1.641 billion by the APR range and amount of the loans that fall in those APR categories. They smartly left off the average term length conditions, so you cannot calculate the present value of those dollar amounts; and instead just have to take the aggregate principal balance future value.
But closer inspection shows that almost all of the securities are at 0% APR… which means something is fishy about the underlying assets. How do you loan customers cars at an APR below 5% if it actually takes a decent amount of interest just to get the money in the first place?
About $1.3 billion of the $1.641 is comprised of loans at 5% or below. Chrysler Financial’s LIBOR rate that they could borrow money is about 5% … so who makes up the difference when ChryslerFinancial lends customers money at a rate worse than what they could have borrowed? You either get it from ChryslerCarCo or you use your own TARP money. Either way, the money used to subsidize ChryslerCarCo and ChryslerFinancial came from the same place.
But to make their game work, money had to come from CarCo so they could actually create a security that was valuable enough to cover what ChryslerFInancial was obliged to pay back. And of course, ChryslerFinancial never has to repay ChryslerCarCo.
It seems a common assumption that CarCo will never pay back their TARP liability. Maybe Chrysler can borrow money from GM so Chrysler can repay its debt so we can keep passing the buck around. Bankers and business guys will always pull a fast one on the public; they win again.
Page S-17 is interesting as well, it shows the % likleyhood that ChryslerFinancial has to go repo a car.
For the investor, the key to this is to make up for the low interest rate by buying the entire security at a discount to par.
They want to sell $1.64 billion in debt for $851.4 million. That works out to be about 52 cents on the dollar. An investor would have to forecast an anticipated repayment schedule that accounts for defaults, repo’s, skips and early prepayment (the last category is a risk on those accounts with high interest rates, because some of them may refi out), and then decide whether 52 cents is a good price.
Now you can see why the domestics can’t make money through volume. When their finance companies lend out money for less than their own cost of capital, then they have to eat losses on the financing. Don’t believe the dealers when they tell you that they’re doing the company a huge favor by selling vehicles with huge incentives on them that produce losses at every turn.
Theoretically, the company might be able to make up for the hit on the spread made on the cars themselves. But since you know that they aren’t offering cut rate financing on popular vehicles, you know that this isn’t going to happen. So this is just a matter of locking in losses.
In this particular case, this may be an easy way to bring in cash at no loss to the current owner, in that I would imagine that these loans were made pre-Fiat; these losses may belong to Cerberus, not so much to Fiat. But that would be a temporary condition, and Fiat is going to find themselves in the same boat soon enough. Once they’ve sold the inventory they’ve inherited and have to pay to build their own product, that’s when it’s going to really start to matter.
I get it. Deciphering with an enigma machine.
Hmmm, I just watched the BBC film FREEFALL last night…it seems to describe exactly what they above are saying about the auto finance industry but about the UK version of the sub-prime debacle…
http://primetime.unrealitytv.co.uk/freefall-launches-on-bbc-one/
Watch it if you can use bbc iplayer, it is a really great (but worrying) film.
http://www.bbc.co.uk/iplayer/episode/b00lrt0p/Freefall/
tl;dr
The difference between the APR offered to consumers and what the APR should be is transferred from the manufacturer to the finance company. Thus, “Buy today and pay 0.0% interest or receive a $1,500 rebate!” If the finance company loses money, it is because 1) defaults are higher than anticipated, 2) they borrow short and lend long, and short rates go up, or 3) the residual value on expiring leases are lower than anticipated.
This marketing money is taken into account when the manufacturer prices the car, just as Proctor and Gamble takes into account its coupons when pricing Tide.
They want to sell $1.64 billion in debt for $851.4 million. That works out to be about 52 cents on the dollar. An investor would have to forecast an anticipated repayment schedule that accounts for defaults, repo’s, skips and early prepayment (the last category is a risk on those accounts with high interest rates, because some of them may refi out), and then decide whether 52 cents is a good price.
What they need to do is securitize/re-tranche that debt, and pay off a rating agency to get all A’s for most of it, sell that part off to suckers for profits/bonuses, and stick the rest onto taxpayers.
Ron has it exactly right. When the captive writes a contract for a promotional rate, the manufacturer writes a check for the difference between the promotional rate and some base rate, after allowing for early terminations (voluntary or involuntary), discounted to present value using some discount rate.
Obviously, both the base rate and the discount rate are key to the amount of subvention paid, are the subject of intense negotiation between the manufacturer and captive, and are closely held secrets. In deference to the captive from which I retired, I’m not going to disclose their formula (it may have changed in the two years or so since I was involved in this stuff).
In any event, Chrysler Financial has cashed the subvention checks for these contracts and is amortizing the subvention into income over the lives of the contracts. Now that the contracts have been securitized, the rest of the subvention sill be brought into income as an offset to the loss on sale of selling a zero percent contract.
The only thing Ron missed is that resale values also affect the profitability of retail contracts because of the effect on the credit loss when you repo a vehicle. So not only are credit losses affected by the default rate, but also by the loss per default. It’s the difference between losing a couple of thousand on an SUV a few years ago and losing ten thousand or more last summer when you couldn’t give the damn things away.
And these are securitizations we’re talking about here (at least with regard to this prospectus). In my experience, the rating agencies kept a close rein on the captives regarding the contracts eligible to be included, tranching, rating of the tranches, etc., so that to the best of my knowledge no external investor has ever lost a penny on a securitization of auto retail contracts from captive finance companies (I won’t comment regarding sub-prime companies). The failure with mortgages was in assuming ever-increasing real-estate prices would bail out the investors, so severities on foreclosures would be effectively zero, so default rate didn’t matter.
In my experience, the rating agencies kept a close rein on the captives regarding the contracts eligible to be included, tranching, rating of the tranches, etc.,
This is true as evident in the prospectus. The majority of the loan recipients have high FICO.
–
The failure with mortgages was in assuming ever-increasing real-estate prices would bail out the investors, so severities on foreclosures would be effectively zero, so default rate didn’t matter.
Not necessarily. One main problem that is practically never reported (because it’s maths, lol) is that by re-tranching, the sensitivity to initial assumptions is very high. It’s a complex situation that can’t/shouldn’t really be summarized to a 1-axis final bonding. Basically, CDO-like things either have to be very conservative, or they’ll exploit a fundamental primitive weakness in bond ratings.