OT: KMX

Hi Jim,

You have mentioned KMX in the past. What do you think about the firm today? Also, are current prices ($92ish) good for backing up the truck?

Thank you for any insights.

I already have more than I need but I’m probably buying more today, if that answers your question.

I think it will be trading over (say) $145 soon enough, and with sufficient certainty, that it will offer a very pleasant return with very low risk.
Very pleasant = double digits, which would have to be by Jan 2027.
My expectation isn’t always the same as reality, but it’s a stake in the ground.

I think some investors don’t really seem to understand their model very well.
It seems to trade based on what people think is happening to used car prices, which matter almost not all to them.
It’s mainly a “flow” business like Visa: they make a relatively constant amount per car, so only volume matters.
That generally goes up when times are good because more people can afford to be buyers.
And it generally goes up when times are tough, since new-car buyers move into the used car space.

The other drumbeat of worry seems to be that the apps without physical locations are going to steal their lunch money.
In short, I don’t buy that. The apps will do OK, but there’s a lot you simply can’t do without physicality.

Jim

15 Likes

Awesome! Thank you very much for the prompt and helpful response, Jim. Much appreciated.

Maybe Buffett should sell some of those crummy auto dealerships that BRK owns in Arizona and put the proceeds into KMX. KMX has a much better reputation for honest dealing and has many customers won’t transact with any other used car dealer.

https://www.carprousa.com/blog/arizona-ag-fines-two-phoenix-…

1 Like

I’ve held a small’ish 2% position of KMX for several years.

Results have been so-so…pretty vanilla but satisfactory to me.

Be prepared its a very lumpy bumpy ride. Mr. Jim is spot on with the idea people buy/sell shares with gut reaction to the unusual used car market past couple years.

Jim, sounds like your conviction is pretty strong. I’ve got a KMX dealership near me and i’ve toured it a couple times. Impressive in its physical size and professional presentation…its a machine. Scuttlebutt with a multi new/used car dealership friend of mine is they tend to ‘pay up’ at dealer auctions in order to keep inventory levels at the churn levels they need, but so far they make it work.

An interesting competitive trend are companies like carbingo.com that will give you a firm price on your used car with a few photos and a brief description from their online app. They seem to really know the market and will write you a check on the spot. They in turn, go to the KMX’s of the world and dealer auction the inventory. Everyone needs a car and dealers needs inventory. KMX needs ‘lots’ of inventory.

For fun, try carbingo.com. You will be followed up a couple times but not overly annoying; they will leave you alone after a couple attempts but you can get a pretty accurate value of your ride.

I also owned AutoNation in tandem with KMX for a while but much prefer the KMX model.

Stick with it, you’ll do fine.

5 Likes

Awesome! Thank you very much for the prompt and helpful response, Jim. Much appreciated.

No problem.
Doesn’t mean I’m right, but you’re welcome to the verbiage.

You should be aware that the “analyst consensus” earnings forecasts do not match mine at all.
For example, the “analysts” expect $7.07 EPS for FY Feb 2024 (mostly calendar 2023), and I expect something closer to $8.50.
Maybe their analysis is much better than mine. If so, the stock isn’t as cheap as I think it is, AND not going to grow as quickly either.

Since there are sometimes one-time hits or gains in the earnings figures, it’s good to use other metrics.
It seems that the P/S ratio is about the cheapest in 12 years, about 30% below the average since 2006.
Since their business does not have a structure wildly different from its recent history, if margins mean revert that’s another data point.
Of course if you think their future has dimmed a whole lot, then the cheapness isn’t a buy signal.
I think they have a good runway yet.

Jim

12 Likes

Be prepared its a very lumpy bumpy ride. Mr. Jim is spot on with the idea people buy/sell shares with gut reaction to the unusual used car market past couple years.

Good point.
I’m a “swing trader” on this. It gets cheap sometimes, and expensive other times.
But the overall results are great, so it also works for a long hold provided you don’t look at the price too often!

Jim, sounds like your conviction is pretty strong.

Now one of my top three positions.
Again, that doesn’t mean I’m right.

An interesting competitive trend are companies like carbingo.com that will give you a firm price on
your used car with a few photos and a brief description from their online app. They seem to really
know the market and will write you a check on the spot. They in turn, go to the KMX’s of the world
and dealer auction the inventory. Everyone needs a car and dealers needs inventory. KMX needs ‘lots’ of inventory.

I get the impression that Carmax is much more of an auction seller than an auction buyer.
They buy a LOT of cars.
Good cars get sold with fixed prices at retail, which is why clients like them: few lemons.
Everything else goes to their weekly wholesale auctions where other dealers buy the second-tier stuff.
That’s the heart of their business model, and the reason they can sell at fixed price:
they know the cars well enough to know with acceptable certainty there aren’t expensive surprises for either party.

For the on line types—Carvana is often mentioned—you mention that they seem to really know the market.
What would worry me if I were a shareholder in one of those companies is the word “seem”.
Pricing bots can sometimes be dumb, sometimes insane. Think of the pricing of mortgages during the credit bubble,
based on reliable statistical techniques applied inappropriately to a too-short too-normal history.
I think the triage process that Carmax does has some real world value, and the algo-only firms have a high risk of stepping into some wet doodoo.

In insurance, I agree that the number crunching types like Progressive with truly better input data (telematics) will have an edge over others.
They know that Bill screeches to a halt at red lights and Bob doesn’t. Geico doesn’t know that.
But I do NOT believe that those using “fancy” programs but comparable data will do better than the dinosaurs at pricing risk over the long run.
It’s very much the same with lenders, and I am betting that it’s the same (or at worst a tie) for car sellers.
That’s because in this case, Carmax has better, not worse, input data than their on line competitors.

Jim

9 Likes

For the on line types—Carvana is often mentioned—you mention that they seem to really know the market

In December 2014, I bought a Honda Fit right off the lot. It had 23 miles, and I paid $20,814.80, including accessories and an extended warranty. This December, a buzzy startup called Carvana drove away with my car, cutting me a check for $20,905 — leaving me with a profit of $90.20.

Not only that, but Carvana’s offer was $5,000 higher than Vroom, $6,000 higher than TrueCar, and $7,500 higher than CarMax. Carvana’s offer changed day by day, too: the final one I accepted was $1,338 higher than its lowest quote.

https://www.theverge.com/22923871/carvana-pandemic-used-car-…

Bot bidding against itself, even when it is the highest bidder, though it may not have known that. The strategy for car sellers might be to get competitive bids and see if one these bots keeps bidding against itself.

2 Likes

This is anecdote, but it might be useful to some.

I hang out with a bunch of younger gearheads most Tuesday nights over pizza and beer. A favorite pastime is scrolling through bring-a-trailer and other car dealer sites on phones, with interesting vehicles for sale passed around as memes. Each of these guys owns upwards of two cars, some own 5 or more. Never once have I heard of them buying a vehicle from any on-line site, except in one case from an obscure estate auction outfit. All their used cars were bought private party via craigslist or word of mouth.

You should be aware that the “analyst consensus” earnings forecasts do not match mine at all.
For example, the “analysts” expect $7.07 EPS for FY Feb 2024 (mostly calendar 2023), and I expect something closer to $8.50.

Jim, I’d be fascinated to hear the process & factors you use to arrive at your own future earnings estimates for companies like this one if you’d be willing to walk through it.

4 Likes

For example, the “analysts” expect $7.07 EPS for FY Feb 2024 (mostly calendar 2023), and I expect something closer to $8.50.

Jim, I’d be fascinated to hear the process & factors you use to arrive at your own future earnings estimates for companies like this one if you’d be willing to walk through it.

Chicken entrails figured prominently.

But really, nothing particularly surprising.
To clarify a bit, it’s not that I expect that specific figure that specific year, but I expect something like that as their normalized earning power that year.
i.e., what that year would come in at if it were neither unusually good nor unusually bad.

The basic thinking starts with the notion that the revenue and revenue growth trends will mostly hold up.
Plus:
We’ll see a resumption of something closer to old net margins as the costs of ramping up their on line offerings fades,
and an unusually good tail wind from buybacks being done at low valuation multiples.
New securitizations should be profitable.

The net margin thing is likely the main reason for the divergence in outlook.
I view the current drop of about 0.6% from prior norms to be transient.
For example, if this year were historically normal on that front, the net would be at least 15% higher.

Or maybe the other analysts believe the end is nigh because of the online-only competitors.
I think that’s perhaps a big part of the reason for the price sell-off, and many analysts tend to follow prices more than business fundamentals.

Another possibility is that I am wildly wrong.

Pure extrapolation isn’t the most solid way to do your reasoning.
But in this case it’s fun, because it certainly gives a cheery view: a twofer.
The rise in earnings per share in the last decade has been a juggernaut.
Build a trend line through those real EPS figures.
(RMS error of the real EPS trend versus actual real EPS is under 5%–a very good trend fit)
Extrapolate the real earnings trend out two years from today.
Imagine that the market price on that day might at be its historical average multiple of trend earnings in the last 10 years = 18.8.
That gives a target price in today’s dollars of $186.80 versus today’s $93. Call it a doubling; we’re among friends.
If that happened, that would be a two year expected return of inflation + 41.7%/year compounded.
That’s not my forecast. But it can be the basis of how to think about things.
To get a result very much worse than that requires a big breaking of either or both of the two inputs:
A very big downward break in the trend real earnings potential and/or a big fall in the market multiples.
e.g., if the earnings trend holds up but the ending market multiple is only 12, you only get inflation + 13.3%/year.
If the two-year-out EPS figures is 2/3 the level implied by the old trend, but the market multiple is its historical usual, you make inflation + 15.7%/year.
If both things happen, you lose -7.5%/year.

Jim

23 Likes

I know little about KMX operations, but we often bumped up against their buyers at the auction. I dreaded when they were interested in the same vehicle I wanted, because they would bid up the car to what I felt was a ridiculous number in order to get it. We were a small volume operation and needed to buy vehicles at a decent value in order to make good grosses, but they didn’t seem to care what they bought it for as long as they bought it.

This was a few years ago before used vehicle prices shot up to ridiculous heights. If KMX was able to navigate the recent used vehicle shortage and price increases fairly painlessly, it seems like they definitely have a model that works.

4 Likes

<We’ll see a resumption of something closer to old net margins as the costs of ramping up their on line offerings fades,

In 2021 calendar year, EPS is already $6.97, and the earning growth is double digit. Should that number be adjusted? Otherwise, it seems low bar to reach $8.5 EPS in two years.

In 2021 calendar year, EPS is already $6.97, and the earning growth is double digit.
Should that number be adjusted? Otherwise, it seems low bar to reach $8.5 EPS in two years.

It does seem like a reasonably low bar. That’s why I’m expecting it.

FWIW, Value Line estimates their “Current EPS” at $7.25 today, in effect what they’re doing right now.
That’s two quarters of history without extraordinary items, plus two quarters of forecasts.
To rise to $8.50 in two years, that would be two years at 8.3%/year increase compounded.
Made up of some mix of revenue growth, share count reduction, and (I hope) a bit of margin expansion.

I think it should be pretty easy to hit $8.50 in nominal terms. (normalized earnings rate two years from now)
It might be harder for them to hit $8.50 in today’s dollars, but probably still not THAT hard.
The trend line through their historical real EPS growth has been inflation + 13.9%/year.
So, they could slow down to 60% of the historical real growth rate and still manage $8.50 in today’s money = inflation + 8.3%/year.

General view:
It is unfortunately necessary to make projections/guesses about top line growth, net margin trajectory, and buyback rate.
And market multiples, of course.
But it seems to me that they don’t have to be particularly optimistic assumptions to come up with a good return from here.

Jim

8 Likes

Would this be one of the DITM option plays that could work out with uncallable leverage?

From a quick glance, the KMX 50 Jan24 CALL is 47.6/48.7 so it has a time value of 6.32, effectively making you borrow 43.68, for an effective 8.2% margin rate with a 1.91 leverage (with 1.77 years to go).

The KMX 45 Jan23 CALL is a bit better, at 48.5/49.5, so 2.12 time value, borrow 42.88, effectively 6.42% margin rate, with a 1.88x leverage (with 0.77 years to go).

Doesn’t look very attractive in terms of margin rate, so I think this one looks like it might be better to simply own the stock.

–Gabriel.

I came to the same conclusion that calls were expensive.

Conversely, selling $95 Jan 23 puts earned me a 16.3% (21.5% annualized) return on the cash backing the put. Or buy it for $82 if put to me.

I know I may end up not owning the stock if it takes off, but I think the 21.5% annualized return is fair trade off.

4 Likes

Would this be one of the DITM option plays that could work out with uncallable leverage?
From a quick glance, the KMX 50 Jan24 CALL is…
I think this one looks like it might be better to simply own the stock.

Indeed–the implied interest rates are pretty high for the Jan 2024 ones.
And Jan 2023 isn’t enough time for a value proposition to work out reliably.

But I suppose calls are suitable if you’re suitably confident.
I (unusually) bought Jan 2023 calls, less than year out, lowest strike.
I might well be wrong, but I’m reasonably confident : )
My thinking for the shorter time frame:
The current implied interest rate for the first not-quite-year (6.4%/yr) is much lower than the current implied rate for the second year (8-11%, depending on the strike).
Pretty good chance of a meaningful bounce this year. You never know, but it seems to make some sense to me.
So, the price will perhaps be much higher when I roll from 2023 to 2024, late this year.
Rolling to any given strike is very much cheaper when the stock price is higher.
So I get the first year of exposure cheaply now, and I hope to get the second year of exposure [more] cheaply later.

As an aside, 6.4%/year is a lot, but it’s less than inflation lately.
So if inflation stays at recent ratess somebody is paying me in real terms to take their money and use it for a year.

Jim

5 Likes

Looking at the cash secured puts on KMX, the nearer term ones seem to offer better rewards than the longer term ones:

May-22 92.5 gives you an entry at 88.4 (if assigned) and 4.64% (58.35 annualized) with a 96.6 FOMO
Jun-22 92.5 gives you an entry at 86.7 (if assigned) and 6.69% (44.67 annualized) with a 98.3 FOMO
Jul-22 92.5 gives you an entry at 85.0 (if assigned) and 8.82% (39.86 annualized) with a 100. FOMO

where the FOMO is what KMX would have to climb to before you were better off having owned the stock directly. Seems rather rich. More gambling rather than investing, but still interesting.

–G

1 Like

not a option related question, but general KMX business related

Does the financial arm of the KMX become an issue during rising interest rate? The default rates are sure to rise as interest rates go up. I am not sure what is the debt on the financial arm of the company but balance sheet shows total debt of $13B. So a percent rise in rates can be meaningful.

But seems like market is not really worried about that much. It seems to be more worried about slowing car sales and its impact on KMX

Any thoughts on losses on the finance side?

3 Likes

Good question, investor. Relatedly, will car sales go down as interest rates rise? (Because financing a car will become more expensive). I guess this would depend on the percentage of cars that are purchased via a loan vs. fully paid outright.