From roughly 2014 through 2021, there was a single piece of car-buying advice you could give almost any client and be right about. Buy used. Finance it cheaply. Take advantage of the depreciation curve someone else has already eaten. Hold for five to seven years. Walk out at the end having driven a meaningfully nicer car for less money than the equivalent new purchase would have cost you. The math was so clean it almost didn't need running.
That math is broken now. Not because used cars got more expensive, though they did, but because financing got more expensive faster, and in a way that disproportionately punishes used-car buyers. I've been running this calculation for friends, readers, and the early Carsmenskii clients I've taken on so far, and the answer keeps coming back surprising. In 2026, depending on the specific car, "buy a year-old certified preowned at the dealer's promotional rate" can be cheaper over a five-year hold than buying the same car privately at $4,000 less and financing through your credit union. That used to be unthinkable. Now it's routine.
The old math (2014–2021)
Quick recap of the world that no longer exists. Between 2014 and 2021, your average enthusiast-leaning buyer in Canada or the US could expect roughly:
- Used-car APR through a credit union: 4.5–6%, depending on credit.
- New-car APR through dealer-arranged financing: 3–4%, often with manufacturer subvention bringing select promotional rates to 0.9–1.9%.
- The premium for "new" over "1-year-old certified preowned" was substantial, typically 18–28% on a $50–$80k car.
Run those numbers and the answer was always the same: skip the new-car premium, buy a year or two old, finance at credit-union rates, and the depreciation cliff someone else paid for becomes your discount. On a $60,000 enthusiast sedan, you'd typically come out roughly $11,000 ahead over five years versus buying the same car new, partly from skipping year-one depreciation, partly from the lower interest costs on a smaller principal.
What actually changed
Two things. The first is obvious; the second is less so.
The obvious one: rates went up. Bank of Canada and Fed rate hikes from 2022 onward pushed prime rates from sub-3% to north of 6%, and credit-union used-car rates followed. As of writing, the typical fair-credit buyer in Toronto is being quoted 8.0–9.5% APR on a 60-month used-car loan. That alone roughly doubles the lifetime interest cost compared to 2019.
The less obvious one: manufacturer captive-finance arms (BMW Financial Services, Porsche Financial Services, Toyota Financial, etc.) responded to rising rates by leaning much harder into subsidized rates on certified preowned inventory. The CPO program, basically "lightly used cars sold by the dealer with a manufacturer warranty extension", used to be a small premium product. In 2026, captives are routinely pushing CPO loans at 4.9%, 3.9%, sometimes 2.9% as a promotional rate, on cars that are 1–3 years old. They do this because the dealer needs to move trade-in inventory, the captive finance arm earns money on the spread regardless of where rates sit, and from the manufacturer's perspective a CPO sale is a customer they retain rather than lose to the secondhand market.
The combination is what broke the old math. An enthusiast buyer who would once have bought private at $50,000 and financed at 5% now faces a choice between buying private at $48,000 and financing at 8.5%, or buying CPO at $54,000 and financing at 4.5%. Run those out over 60 months, and the second option costs less over the life of the loan. The dealer markup is more than offset by the rate differential.
Why CPO is winning right now
Let me run an actual scenario. Two cars, both 2022 BMW M340i xDrive, both 35,000 km, both clean. One is a private sale at $52,000. The other is at a BMW dealer as CPO at $58,500, with a 24-month warranty extension and access to BMW Financial's promotional rate.
The private buyer puts $10,000 down, finances $42,000 at 8.5% over 60 months. Monthly payment: ~$862. Total interest paid: ~$9,720. Total cost over 60 months including down payment: $61,720.
The CPO buyer puts the same $10,000 down, finances $48,500 at BMW Financial's 4.49% promotional rate over 60 months. Monthly payment: ~$903. Total interest paid: ~$5,680. Total cost over 60 months including down payment: $64,180.
So the private buyer "wins" the comparison by $2,460. But – and this is the part most buyers don't think about, the CPO car came with 24 months of additional manufacturer warranty, which on this generation of M340i typically covers $2,000–$3,500 of expected repair work over those two years. The private buyer is exposed to that repair risk. And I'm holding rate constant; if rates trend back down before the loan matures (likely), the CPO buyer's option to refinance is also somewhat better because BMW Financial historically offers refi promotions on existing loans.
Net adjusted cost difference over five years: roughly $0–$1,000 between the two options, depending on your repair luck. Versus a 2019 version of this same comparison, where the private buy would have come in $4,500–$6,000 cheaper. The gap closed because rates moved more than purchase prices did.
The new buy-used rules
Here's what I'm now telling clients and Carsmenskii readers. None of this is universal, every car and every buyer is different, but the principles have flipped enough that it's worth working through carefully.
1. If you're financing, run the CPO comparison every single time. Never assume private is cheaper. Get the dealer's actual quoted CPO rate (not their advertised one), get your credit union's actual quoted rate (not their advertised one), and run both five-year totals. About 60% of the time in 2026, CPO wins. In 2019 that number was closer to 10%.
2. If you're paying cash, the calculation flips back to the old version. No financing means no interest-rate sensitivity, which means depreciation is the only thing that matters, which means private (or auction) is almost always cheaper. The cash buyer is now the only buyer for whom the pre-2022 advice still holds cleanly.
3. Shorten the loan term, not the price. If you must finance, a 36-month loan at 8.5% costs less in total interest than a 72-month loan at the same rate, even though the monthly is higher. In a high-rate environment, the buyer who can stomach a higher monthly payment over a shorter term genuinely beats the buyer who stretches to keep monthlies low.
4. Be aggressive about putting more down. Every additional $1,000 of down payment saves about $260 of interest at current rates over a 60-month loan. The down payment is the cheapest cost reduction available to the financed buyer.
What I'm telling clients
The advice I now give every matchmaking client at the start of an engagement, before I even start the search:
"Tell me how you're financing this. The answer changes which cars I'm looking at, which dealers I'm calling, and what your real budget is."
That used to be a tertiary question. Now it's the second question, after "what kind of car." A buyer paying cash and a buyer financing at 8.5% and a buyer who has access to a 4.5% CPO promo are three different clients shopping three different markets, even if they all walk in saying "I want a 2-year-old M340i for $55,000." Two years ago the answer would have been the same private listing for all three of them. Today the answer is three different cars.
This is the kind of structural change that quietly reshapes the entire used-car market over a decade and that most buyers, who buy a car maybe once every 4–7 years, completely miss. The dealership that sold them their last car in 2019 told them a true story. The same dealership in 2026 is telling them a different story, and the differences are easy to dismiss as "marketing" when they're actually structural.
When this advice expires
One honest caveat. The whole CPO-wins calculation depends on captive lenders continuing to subsidize their promotional rates, and on private-market rates staying high. Both of those are policy variables. If the Bank of Canada starts cutting aggressively (some forecasters expect a cut path through late 2026), the spread compresses and private-buy logic starts to recover. If captives pull back on promotional rates because their financing portfolios are getting too concentrated in CPO, same thing.
So the advice in this essay is unambiguously correct in April 2026. By April 2027, some of the math may have moved. The principle, always run the comparison, don't assume private is cheaper, is permanent. The specific dollar advantages I'm citing are not.
If you're in the middle of a buying decision and the math feels off, it probably is. Book a call and I'll run the actual numbers on the actual car you're considering. It's the kind of problem that takes 20 minutes to clarify and saves $3,000 over the next five years.
