Credit Acceptance: 2020 Collections Wrap-Up

Summary

• Latest collections performance.
• Updated forecasts for IRR by loan vintage.
• A word on repurchase activity.

The purpose of this note is to provide an update on experienced as well as modeled loan/collections performance at Credit Acceptance (NASDAQ:CACC) based on the company's most recent (2020 year-end) update.

For comparison, the last update I provided is over a quarter stale. For the reader new to this series of updates (or for a refresher on the methodology being used and context), I suggest perusing the inception article of the series.

Regardless, a brief reminder: for 'realized IRR' I am effectively providing IRR calculations based on actual cash in/out data provided by the company. This is a backwards looking exercise.

The `modeled IRR' is a forward looking exercise that attempts to forecast eventual IRRs based on historic delinquency curves and the most recent collections data available. Crucially, no adjustments are made based on macro-economic factors.

As usual, we start with IRRs already realized for well-seasoned vintages:

Year	Expected collections realized	Realized IRR
2011	99.8%	26.4%
2012	99.7%	24.4%
2013	99.4%	22%
2014	98.9%	20.4%
2015	97.8%	18.2%
2016	93.8%	15.6%
2017	84.3%	12%

Click to enlarge

Note that CACC's eventual IRRs on the above vintages will be better than the stated amounts, since the calculations above only take into account actual cash flows that have already been received. Of course, this improvement will be minor for 2015 and prior vintages, since the vast majority of the expected collections have already been received for these.

Next up, we have the meat of this exercise - modeled IRRs. For comparison, I have provided my initial forecasts as well as the forecast from my last update (a track record, if you will):

	Expected collections realized	Initial forecast	Prev. forecast (Nov 2020)	Current forecast
2011	99.8%	26.40%	26.40%	26.4%
2012	99.6%	24.50%	24.40%	24.4%
2013	99.3%	22.20%	22.20%	22.1%
2014	98.8%	20.30%	20.80%	20.6%
2015	97.3%	18.7%-20.3%	18.7%	18.9%
2016	92.2%	17.2%-20%	17%	17.4%
2017	81.3%	16.4%-20%	16.3%-16.7%	17%
2018	63%	14.2%-20%	13.8%-14.5%	14%-14.5%
2019	38%	13%-17%	12%-14%	12.5%-13.5%
2020	11.1%	12%-15%	12%-15%	12%-14%

Click to enlarge

I would summarize this data as follows. The 2011-2017 vintages are more or less in the clear as far as profitability is concerned. The 2018-2020 vintages are trending towards the lower end of the forecast range. From the mechanical perspective of the model there don't seem to be any major surprises.

The usual caveat is in order at this point: the projected IRRs overstate the returns accruing to the company, since they do not account for significant collection costs that are part and parcel of the business model (one should lop off about 5-7% from the above numbers to account for these). On the other hand, these IRRs understate returns to equity due to the presence of leverage. Finally, all of these returns are pre-tax.

Regardless, the economics of the loans the company is making still hold up. However, one should not expect profitability at the levels of the 2016 and prior era.

A word on repurchase activity

Credit Acceptance tends to approach share repurchases opportunistically (as opposed to an always on mechanical program). So the following activity is noteworthy.

During 2020 the company repurchased approximately 5.5% of the shares outstanding for an average cost of roughly $375 per share. At pixel, the shares are trading at about $365 per share.

I invite the reader to draw their own conclusions from all of this.

Analyst’s Disclosure: I am/we are long CACC. I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.