Welcome to Tesla Motors Club
Discuss Tesla's Model S, Model 3, Model X, Model Y, Cybertruck, Roadster and More.
Register

Tesla, TSLA & the Investment World: the Perpetual Investors' Roundtable

This site may earn commission on affiliate links.
That data is problematic due to the multiple invoices per car issue. They divide by two to get the real number but some posters report more than 2 invoices in their accounts for a single car. Especially after the EAP->AP dust was settling.

I agree that the absolute numbers are problematic.

Nevertheless the time sorted series looks useful to determine take-rate developments: the take-rate trends I outlined are consistent trends with little noise, and that kind of conclusion wouldn't be sensitive to the sampling rate of the invoices which changed during the quarter.

I.e. I think the following trends are probably real:

  • Very interesting trend regarding FSD take-rate: in the last 1 month of orders FSD take-rate has increased from 10% to almost 40% (!), and with another 45% of Autopilot orders the combined take-rate is now up to 85%. Only 15% of customers didn't order Autopilot. This is a massive shift in FSD take-rate.
  • "Solid black" percentage increased from 25% early orders to 35% in later orders: this suggests that some budget constrained customers "saved" money by not paying $1,500 for color, but opted for the $3k+$5k AP and FSD instead. (The other explanation is that black car color is useful in northern countries such as Norway, as it would capture heat in the winter, snow/ice would melt faster when the car is parked outside, etc.)
  • ASP remained around the €63k-€65k range and remained centered around €64k, which suggests that the €3,100 + €5,200 AP and EAP options are "financed" by customers deselecting other options such as color or white interior. This increases margins and speeds up production in Fremont, which is a win-win for Tesla.
  • It's the first time we can see this phenomenon of customers shifting towards 100% margin software options in actual customer order data.
  • I.e. this is very good news for margins.

Part of it is consumers ordering FSD to guarantee a free HW3 upgrade; but it's also the physical car as a software platform, where budget constrained customers consciously down-scale physical features to pay for software features like AutoPilot and FSD, showing the first signs of life. :D
 
New Baird rating came out yesterday, reiterating "Outperform":

Baird analyst Ben Kallo reiterated an Outperform rating and $465 price target on Tesla and would buy stock ahead of the Q1 delivery announcement likely before 4/3. The analyst thinks investors will look through the noise of the ITC cut [Q4 tax credit cliff] and focus on international deliveries.

The analyst stated "Model 3 deliveries will likely be sequentially softer in Q1 as the delivery focus shifted internationally (we expect elevated cars in the channel), but we think this phenomenon is well understood at this point. For Q1, we model 17k Model S+X deliveries and 50k Model 3, vs. FactSet consensus of 19.7k/55.5k. That said, we think the market expectation for Model 3 deliveries is closer to 50k and believe the softer deliveries are likely priced in at current levels, particularly as sell-side estimates catch-up".​

I've highlighted the 2019/03/28 snapshot of the FactSet consensus:

19.7k S+X deliveries
55.5k Model 3 deliveries
= 75.2k total deliveries​

I believe the S+X deliveries are optimistic - they'll probably in the 15k-20k range, with a real possibility of it being in the lower end of the range.

in 2018 Q1 S+X deliveries were 21,800, and Tesla guided in Q1'19:

"Because this high level of demand presumably represented a pull-forward, we are expecting our Model S and Model X deliveries in Q1 2019 to be slightly below Q1 2018."​

Note that the FactSet "consensus" is a bit below 20k, making it probable that Tesla will "miss" both the consensus and their own guidance. I also guess that S+X deliveries below 20k will be reported as a "big miss" by the usual suspects - so even if the delivery report is actually good overall, Wall Street data services and news agencies reporting and the resulting price action might not match it.

Also note how Baird is pointing it out that for Model 3 deliveries the market expectations are actually around the 50k deliveries range - yet FactSet is over-estimating it with 55.5k deliveries, making a "miss" possible.

We've seen these kinds of FactSet driven "Wall Street expectations" market manipulation shenanigans a couple of times in the past, so watch out for this possibility and be careful with any leverage.
 
New Baird rating came out yesterday, reiterating "Outperform":

Baird analyst Ben Kallo reiterated an Outperform rating and $465 price target on Tesla and would buy stock ahead of the Q1 delivery announcement likely before 4/3. The analyst thinks investors will look through the noise of the ITC cut [Q4 tax credit cliff] and focus on international deliveries.

The analyst stated "Model 3 deliveries will likely be sequentially softer in Q1 as the delivery focus shifted internationally (we expect elevated cars in the channel), but we think this phenomenon is well understood at this point. For Q1, we model 17k Model S+X deliveries and 50k Model 3, vs. FactSet consensus of 19.7k/55.5k. That said, we think the market expectation for Model 3 deliveries is closer to 50k and believe the softer deliveries are likely priced in at current levels, particularly as sell-side estimates catch-up".​

I've highlighted the 2019/03/28 snapshot of the FactSet consensus:

19.7k S+X deliveries
55.5k Model 3 deliveries
= 75.2k total deliveries​

I believe the S+X deliveries are optimistic - they'll probably in the 15k-20k range, with a real possibility of it being in the lower end of the range.

in 2018 Q1 S+X deliveries were 21,800, and Tesla guided in Q1'19:

"Because this high level of demand presumably represented a pull-forward, we are expecting our Model S and Model X deliveries in Q1 2019 to be slightly below Q1 2018."​

Note that the FactSet "consensus" is a bit below 20k, making it probable that Tesla will "miss" both the consensus and their own guidance. I also guess that S+X deliveries below 20k will be reported as a "big miss" by the usual suspects - so even if the delivery report is actually good overall, Wall Street data services and news agencies reporting and the resulting price action might not match it.

Also note how Baird is pointing it out that for Model 3 deliveries the market expectations are actually around the 50k deliveries range - yet FactSet is over-estimating it with 55.5k deliveries, making a "miss" possible.

We've seen these kinds of FactSet driven "Wall Street expectations" market manipulation shenanigans a couple of times in the past, so watch out for this possibility and be careful with any leverage.

Market should still be happy if S+X misses consensus by a few thousand but Model 3 beats it by ten thousand, or whatnot.
 
Let's have a look at probable "Tesla Q1 Vehicle Production and Deliveries" report filing dates.

Firstly, here's the Edgar filing track record and timestamps of D&P reports from the past 3 years:

Code:
2016/Q1: Filing Date: 2016-04-04 (Mon), Accepted: 2016-04-04 16:57:12 (Mon), Period of Report: 2016-04-04 (Mon)
2016/Q2: Filing Date: 2016-07-05 (Tue), Accepted: 2016-07-05 06:17:10 (Mon), Period of Report: 2016-07-03 (Sun)
2016/Q3: Filing Date: 2016-10-03 (Sun), Accepted: 2016-10-03 16:47:46 (Sun), Period of Report: 2016-10-02 (Sat)
2016/Q4: Filing Date: 2017-01-03 (Tue), Accepted: 2017-01-03 16:25:20 (Tue), Period of Report: 2017-01-03 (Tue)
2017/Q1: Filing Date: 2017-04-03 (Sun), Accepted: 2017-04-03 06:02:22 (Sun), Period of Report: 2017-04-02 (Sat)
2017/Q2: Filing Date: 2017-07-03 (Mon), Accepted: 2017-07-03 15:21:19 (Mon), Period of Report: 2017-07-03 (Mon)
2017/Q3: Filing Date: 2017-10-02 (Mon), Accepted: 2017-10-02 17:19:16 (Mon), Period of Report: 2017-10-02 (Mon)
2017/Q4: Filing Date: 2018-01-03 (Wed), Accepted: 2018-01-03 17:10:27 (Wed), Period of Report: 2018-01-03 (Wed)
2018/Q1: Filing Date: 2018-04-03 (Tue), Accepted: 2018-04-03 09:00:33 (Tue), Period of Report: 2018-04-03 (Tue)
2018/Q2: Filing Date: 2018-07-02 (Mon), Accepted: 2018-07-02 09:04:06 (Mon), Period of Report: 2018-07-02 (Mon)
2018/Q3: Filing Date: 2018-10-02 (Tue), Accepted: 2018-10-02 08:35:28 (Tue), Period of Report: 2018-10-02 (Tue)
2018/Q4: Filing Date: 2019-01-02 (Wed), Accepted: 2019-01-02 08:38:32 (Wed), Period of Report: 2019-01-02 (Wed)

These are the Q1 timestamps only:

Code:
2016/Q1: Filing Date: 2016-04-04 (Mon), Accepted: 2016-04-04 16:57:12 (Mon), Period of Report: 2016-04-04 (Mon)
2017/Q1: Filing Date: 2017-04-03 (Sun), Accepted: 2017-04-03 06:02:22 (Sun), Period of Report: 2017-04-02 (Sat)
2018/Q1: Filing Date: 2018-04-03 (Tue), Accepted: 2018-04-03 09:00:33 (Tue), Period of Report: 2018-04-03 (Tue)

From this the historic pattern is:
  • Tesla filed on either April 3rd or 4th, and it's weekend-invariant such as in Q1 2017.
  • The last 2 years Tesla filed on the 3rd.
  • But note that the last 3 quarterly P&D reports are showing a speedup in reporting on the 2nd day of the next month - i.e. 2019 could see a Q1 report date on April 2, Tuesday.
  • The timestamps are Eastern Time, and the last 4 reports were filed between 8:30am and 9:15am, just before regular trading begins at 9:30am.
  • In earlier years Tesla filed the report after the close of trading, with one exception (Q2'17).
From this my expectation are the following dates and probabilities:

The Q1 deliveries and production report will probably be released on one of the following dates:
  • April 2 (Tuesday), before market open on a Wednesday, between 8:30am and 9:30am. I'd say the probability is ~80% as this is the latest pattern IMHO.
  • April 3, before market open on a Thursday. Probability ~20%: this is what they did in 2018, but then sped up their reporting system in the rest of 2018.
  • There's tail probabilities of late April 1 release (unlikely due to the 'April's fool' date), or after regular trading on the 2rd, or anytime on the 3rd. Any other dates would set a new precedent.
  • Note that Wall Street appears to be expecting April 3rd based on the 2018 precedent, so an "early" report released on the 2nd before trading could create stronger price reaction than usual.
I.e. including today there's probably 2 trading days left (March 29 and April 1) before the delivery report is released.

Not advice. :D
 
Last edited:
Market should still be happy if S+X misses consensus by a few thousand but Model 3 beats it by ten thousand, or whatnot.

S+X deliveries matter to perception: in Q4 half of Tesla's cash generation came from the S+X, the other half from the Model 3.

Margins are improving rapidly on both, so I agree with you that even 15k S+X deliveries would still contribute a nice chunk of cash to Q1 financials if Model 3 deliveries are better, but I think the usual bear raid on any excuse of a technical "missed Wall Street expectations" false narrative is possible, in an attempt to nullify the positives in the Q1 production report.

Anyway, it doesn't have to happen that way, but:
  • The Q4 delivery report signaled a well executed bear raid as well,
  • the Q3 delivery numbers were actually lied about reported erroneously on the financial news wires that most institutional investors are reading ... It was "corrected" later in the day, but by that time the damage was done and the bad figures and conclusions were syndicated into hundreds of secondary articles (most of which never got corrected/updated). The lasting perception in most investors was that Tesla "missed", despite having met both guidance and exceeding FactSet expectations.
So be careful - this is TSLA. ;)
 
Last edited:
Thanks for the data point - but on a large scale, this is just noise. I was called to pick up my Model 3 earlier, and have read other stories of both earlier and later deliveries. Tesla said they would start delivering cars in mid February in Europe and to everyone´s surprise here, the first car in Germany was delivered on 2/13 (sourc: Das erste Kunden Model 3 in Deutschland ist gelandet! • TFF Forum - Tesla Fahrer & Freunde even a day or two earlier in the Netherlands). So while individual appointments have been a mess, from what I observed this was not because ships were late etc but because Tesla didn´t know where the cars were and internal communication problems.

TLDR: Yes, there were individual delays, but nothing that would impact delivery numbers to more than a few percent IMHO.
On that note, not sure if you`ve seen Alex`s latest video... It is always so painful to listen to these stores. So stupid.

One would think that Tesla could hire some experienced customer service supervisors who could set up the processes. This is not rocket science. At one point the team I managed was responsible for 120.000 customer (IT) requests per year and I got to handle the escalations and the VIP complaints. You just need to collect the data, drill down and implement simple methods, like weekly reviews of the longest outstanding items, or most common complaints, automate as much of the standard processes as possible and keep pushing until you see the improvements. This can be done and there are thousands of people out there with experience in this.

 
It should be reiterated that S&P inclusion is not an automatic process. That said, so long as Tesla unambiguously meets the guidelines, I'd think it hard to reject them.
Why would you think that? We know Moody's isn't going to upgrade their credit rating until it is beyond ridiculous because the shorts (Wall Street generally) has their ear. I see no reason to believe that Standard and Poor is any different. I think TSLA could meet S&P guidelines for two straight years without being included because they too are a part of the Wall St. machine that wants to destroy Tesla. We could hope it's different but I certainly wouldn't expect it to be.
 
Market should still be happy if S+X misses consensus by a few thousand but Model 3 beats it by ten thousand, or whatnot.
I am with @Fact Checking on this one.

Can I get you a taste of the April 2/3 headlines?

Tesla Model S & X sales crumble as Model 3 deliveries fall, raising questions on demand and Tesla`s cash crunch

Model 3 cannibalizes Model S/X sales while falling short of Wall Street expectations

Disappointing demand for Tesla`s vehicles validates the end of the growth story


Need more? I can go on all day long.

No matter what records will be broken, what the company itself has guided for, it is easy to cling to 1 lower data point and manufacture a crisis. The evidence is already in with the EPS "consensus" for Q1 even though for 6 months they have been saying Q1 will be break even at best and Elon has now specifically guided for a loss. Yet, even if they now post a surprise profit, there is still a chance it will be framed as a miss.

From Yahoo Finance:
upload_2019-3-29_11-37-6.png
 
JFYI, the pre-market price drop is most likely due to JP Morgan analyst Ryan Brinkman "downgrading" Tesla:

JPMorgan analyst, Ryan Brinkman, reiterated his Underweight rating on shares of Tesla and cut his price target to $215 (from $230) after taking Martin Viecha, Director of Investor Relations, to meet with investors.

Key points from the meetings include:

The $35K standard version Model 3 will generate substantially lower margin than higher priced variants and contribute only a slightly positive variable cash. Tesla's ability to meet Model 3 margin and cash targets seems to be highly dependent on the take rates for its high-margin $3,000 and $8,000 Autopilot packages.

There have been some delays in getting Model 3s to customers in 1Q. A since resolved, customs labeling issue in China and a dockworkers strike in Europe are two key examples.

Model 3 ASPs are likely to decline sequentially in 1Q as the headwind of a higher mix of lower spec Model 3s for the US market will more than offset the tailwind of shipping high-ASP AWD and performance versions to Europe and China.​

Notes:
  • The $230 price target was already out of whack with financial multiples - the $215 is just a bit more ridiculous
  • The FSD package is not $8,000 but $5,000. The weird wording makes it appear as if customers would have to pay $11,000, while the real maximum cost is $8,000.
  • Brinkman's fear-mongering about AutoPilot take-rates is unfounded: AutoPilot take rates are over 90% in European data, FSD take-rate up to 40%.
  • As I estimated it before even the $35k generates positive cash (!) - which Brinkman tries to obfuscate as "contributes only slightly positive to variable cash".
  • Brinkman also missed that fact that Tesla has a separate 6-8 weeks long waiting list for the $35k SR version, and it's entirely up to Tesla how many base versions they are selling. Tesla can rate-limit SR production as long as margins are not good enough yet.
  • That the $35k car generates cash is actually very good news, because both SR+, color and AutoPilot/FSD take-rates are very high according to various trackers:
    • SR+ take-rate is higher than 70%,
    • non-black color take rate is around 75%,
    • AP take-rate is above 80% in the U.S., above 90% in Europe,
    • FSD take-rate has reached 40% in recent European orders.
  • Potential conflicts of interest were not disclosed by Brinkman: it's unclear how large JP Morgan's and their biggest clients' net short position in TSLA is, in the shares and stock options markets.
Ryan Brinkman's TSLA ratings track record on TipRanks is awful: -10% returns, many Tesla downgrades given near lows. He has issued over 20 "sell" ratings for Tesla in the past. It seems obvious that Brinkman is just talking JP Morgan's book and is taking one for the short team ...

TL;DR:
JP Morgan's Brinkman is writing nonsense about Tesla and is actively misleading about the Model 3 Standard Range's margins and take-rates.
 
Last edited:
I am with @Fact Checking on this one.

Can I get you a taste of the April 2/3 headlines?

Tesla Model S & X sales crumble as Model 3 deliveries fall, raising questions on demand and Tesla`s cash crunch

Model 3 cannibalizes Model S/X sales while falling short of Wall Street expectations

Disappointing demand for Tesla`s vehicles validates the end of the growth story


Need more? I can go on all day long.

No matter what records will be broken, what the company itself has guided for, it is easy to cling to 1 lower data point and manufacture a crisis. The evidence is already in with the EPS "consensus" for Q1 even though for 6 months they have been saying Q1 will be break even at best and Elon has now specifically guided for a loss. Yet, even if they now post a surprise profit, there is still a chance it will be framed as a miss.

From Yahoo Finance:
View attachment 391471
Of course it will! This is Tesla we're talking about after all. You forget the first principal in the financial analyst playbook. "No good financial quarter for Tesla shall go unpunished." Did you miss that day in Financial Analyst 101 class? lol!

Dan
 
I am with @Fact Checking on this one.

Can I get you a taste of the April 2/3 headlines?

Tesla Model S & X sales crumble as Model 3 deliveries fall, raising questions on demand and Tesla`s cash crunch

Model 3 cannibalizes Model S/X sales while falling short of Wall Street expectations

Disappointing demand for Tesla`s vehicles validates the end of the growth story


Need more? I can go on all day long.

No matter what records will be broken, what the company itself has guided for, it is easy to cling to 1 lower data point and manufacture a crisis. The evidence is already in with the EPS "consensus" for Q1 even though for 6 months they have been saying Q1 will be break even at best and Elon has now specifically guided for a loss. Yet, even if they now post a surprise profit, there is still a chance it will be framed as a miss.

From Yahoo Finance:
View attachment 391471

If we were talking about a small Model 3 beat, and a small S+X miss, I would agree with that sentiment. But if it's a small S+X miss, but an utter Model 3 blowout, the latter part simply can't be ignored. It crushes the entire "Model 3 demand hypothesis", which is one of the key worries about the stock at present (the other being the SEC case). So long as the Model 3 cash cow keeps mooing, the company is fine.
 
Not worth millions but just looking at the company historically, it hovered between 4-6 dollars for years and only dropped to 2 the past 7-8 months, which means a lot of longs are getting burned by this deal.

No burned by the deal. They bought a stock that tanked. The acquisition offer nearly doubled the price, that is good for them. A stock exchange would mean no tax consequences (stupid IRA wash sale issues) which is favorable. Then they need to look at who has a higher chance of growth going forward, TSLA with price targets of >50% current price, or MXWL which has apparently been plodding along at 4-6 for years.

The burn already happened, this is an opportunity to switch bodies.
 
If we were talking about a small Model 3 beat, and a small S+X miss, I would agree with that sentiment. But if it's a small S+X miss, but an utter Model 3 blowout, the latter part simply can't be ignored. It crushes the entire "Model 3 demand hypothesis", which is one of the key worries about the stock at present (the other being the SEC case). So long as the Model 3 cash cow keeps mooing, the company is fine.

See today's JP Morgan 'downgrade': the shortz are already pivoting their Q1 arguments, if there's an upside surprise to deliveries they will remain "worried about Model 3 margins".

And since delivery reports usually do not disclose financials, they will remain 'worried' until the Q1 earnings report, at which point they'll switch on to be "worried" about Q2, etc. :D

So I'd caution against leveraged optimism related to the deliveries report: the dice are loaded, the kicks are below the belt and Tesla's deck of cards has all the aces removed. ;)
 
(The most efficient gasoline car in existence only gets 58 mpg -- and that's a non-plug-in hybrid. The most efficient non-hybrid gets less than 40 mpg, last I checked -- I can't find them on the "top" lists any more)

Per the latest version of the 2019 Fuel Economy guide, the most fuel efficient vehicle of each type:

TypeModelMPG combined, unrounded, adjustedMPG combined, unrounded, unadjusted
PHEV, charge depletingToyota Prius Prime133.2661 MPGe190.3803 MPGe
BEVTesla Model 3 Long Range (RWD)130.2464 MPGe175.182 MPGe
FCVToyota Mirai67.2229 MPGe96.0327 MPGe
Non-plug-in hybridHyundai Ioniq Blue57.782477.1931
PHEV, charge sustainingToyota Prius Prime54.432976.4075
GasolineMitsubishi Mirage38.547854.441
DieselChevrolet Cruze36.8031 (32.9397 MPGe)50.0238 (44.7726 MPGe)

I'm using unrounded but adjusted numbers in place of the window sticker numbers, because why throw away the accuracy that the EPA data has?

Unadjusted numbers are the numbers that CAFE uses, with the caveat that some advanced technology vehicles (not sure about fuel cells, but PHEVs in charge depleting mode without fuel usage, and BEVs) are currently counted as infinite MPG, and even with modifiers making them count as multiple cars.

And, diesel MPGe is calculated by MPG / (128,488 / 115,000), based on AFDC data on ultra low sulfur diesel and gasoline energy densities. (IIRC CAFE actually works off of CO2 on the back end, but the difference between the fuels is pretty similar.)

Did you get the Cambria email? I’m curious what the actual process is for tendering shares.

Granted, I bought MXWL after this all started, but I have not gotten anything from Cambria. I need to contact Robinhood and see what my options for tendering are...
 
Last edited:
Stock drop premarket - another downgrade?

Yeah, see the JP Morgan downgrade from $230 to $215 (!) with hilarious concern trolling by analyst Ryan Brinkman:

JFYI, the pre-market price drop is most likely due to JP Morgan analyst Ryan Brinkman "downgrading" Tesla:

JPMorgan analyst, Ryan Brinkman, reiterated his Underweight rating on shares of Tesla and cut his price target to $215 (from $230) after taking Martin Viecha, Director of Investor Relations, to meet with investors.

Key points from the meetings include:

The $35K standard version Model 3 will generate substantially lower margin than higher priced variants and contribute only a slightly positive variable cash. Tesla's ability to meet Model 3 margin and cash targets seems to be highly dependent on the take rates for its high-margin $3,000 and $8,000 Autopilot packages.

There have been some delays in getting Model 3s to customers in 1Q. A since resolved, customs labeling issue in China and a dockworkers strike in Europe are two key examples.

Model 3 ASPs are likely to decline sequentially in 1Q as the headwind of a higher mix of lower spec Model 3s for the US market will more than offset the tailwind of shipping high-ASP AWD and performance versions to Europe and China.​

Notes:
  • The $230 price target was already out of whack with financial multiples - the $215 is just a bit more ridiculous
  • The FSD package is not $8,000 but $5,000. The weird wording makes it appear as if customers would have to pay $11,000, while the real maximum cost is $8,000.
  • Brinkman's fear-mongering about AutoPilot take-rates is unfounded: AutoPilot take rates are over 90% in European data, FSD take-rate up to 40%.
  • As I estimated it before even the $35k generates positive cash (!) - which Brinkman tries to obfuscate as "contributes only slightly positive to variable cash".
  • Brinkman also missed that fact that Tesla has a separate 6-8 weeks long waiting list for the $35k SR version, and it's entirely up to Tesla how many base versions they are selling. Tesla can rate-limit SR production as long as margins are not good enough yet.
  • That the $35k car generates cash is actually very good news, because both SR+, color and AutoPilot/FSD take-rates are very high according to various trackers:
    • SR+ take-rate is higher than 70%,
    • non-black color take rate is around 75%,
    • AP take-rate is above 80% in the U.S., above 90% in Europe,
    • FSD take-rate has reached 40% in recent European orders.
  • Potential conflicts of interest were not disclosed by Brinkman: it's unclear how large JP Morgan's and their biggest clients' net short position in TSLA is, in the shares and stock options markets.
Ryan Brinkman's TSLA ratings track record on TipRanks is awful: -10% returns, many Tesla downgrades given near lows. He has issued over 20 "sell" ratings for Tesla in the past. It's obvious that Brinkman is just talking JP Morgan's book and is taking one for the short team ...

TL;DR: JP Morgan's Brinkman is writing nonsense about Tesla and is actively misleading about the Model 3 Standard Range's margins and take-rates.

What I don't understand, why is Martin Viecha, Tesla Director of Investor Relations wasting time meeting with dishonest short side analysts like JP Morgan's Brinkman?

Is there no end of quarter delivery push that needs volunteers, or something? ;)
 
Last edited:
That's what I call a missed referral opportunity.
A referral opportunity might have backfired because it implies you are getting something if he purchases the car. If the person is a real friend, it's different because they know that's not the motivation. Someone who's just an acquaintance does not.