Driverless Cars

Tesla’s FY2017 – cashflow stunts bigger than a roadster in orbit


No beating around the bush. Tesla’s cash-flow situation resembles that of One.Tel in Australia before it became insolvent. Rocketing financing and investing cash-flow with troubled operating cash which in Tesla’s case was flattered by some accounting trickery.  The Q4 2017 earnings release spoke of fairies and magic pixie dust for the most part. Q1 deliveries to date look to undershoot.  Once again a promise to hit production of 2,500 Tesla Model 3s by the end of Q1 and 5,000 a week by end of Q2 2018 (i.e. 6 months away). Note that Tesla had about 860 undelivered Model 3 cars at the end of Q4. That is a high ratio given 1550 were shipped in Q4.

While the company claims a cash balance of $3.4bn which many will pop champagne corks over, Tesla has accrued liabilities, accounts payable and customer deposits totaling $4.975bn at quarter end. This also excludes the $608mn in extra ‘residual value guarantees’ on the books YoY.

The company expects to break even during the year. However with gross automotive margins about to suck up the Model 3 in larger numbers that will take some doing despite claims it can do 25% vs the existing line-up’s 18% range. As at January, Q1 sales in the US are at 2016 levels and European registrations are down around 14% in aggregate across Norway, Holland, Italy, Belgium, Sweden, Austria and Switzerland. Lots can change but it doesn’t read well to kick off 2018’s challenge to break even at an operating level. The Model 3 is on average two-thirds cheaper than the average selling price on existing products so to even hold margins constant will take the mother of all cost cutting all the meanwhile facing new competition over 2018 which will weigh on pricing.

Interesting within the operating cash-flow statement is a term “Changes in operating assets and liabilities,net of effect of business combinations” which shows a quarter on quarter swing of $746.8m pushing net operating cash to +$509mn achieving a new quarterly record. This was achieved mainly by improved collection of receivables (believable), inventory reduction of finished vehicles (were incomplete vehicles that left the factory to parking lots yet to be delivered due to a lack of parts counted?), improved working capital from the ramp of Model 3, and growth in customer deposits (this was only  $168m QoQ vs expectations of $400m) from Semi and Roadsters that were announced with fanfare during Q4. Cash burn appeared lower because the company included customer deposits for the upcoming Semi and Roadster in its operating CF. That is slightly deceiving because deposits aren’t supposed to be drawn from current operations. The Roadster is supposed to be ready by 2020. This seems odd.

Tesla wrote “Despite the delays that we experienced in our production ramp, Model 3 net reservations remained stable in Q4.” Strange there was no mention of progress on Roadster and Semi orders in Q4. Was the $250,000 deposit within 10 days for the Founder series Roadster a bit steep? Truck orders seem around 600-700 at this stage and at $5,000 a deposit, generously speaking $3,500,000 isn’t a swing. As mentioned earlier the +$168m in customer deposits could only reflect how poorly orders for those vehicles are tracking such is the need to avoid talking about them in the statement (surely something to crow about) other than projected performance stats.

Capital expenditures in 2018 are projected to be slightly more than 2017 according to the statement. Tesla also mentioned “quarterly operating income should turn sustainably positive at some point in 2018.” That is a hugely optimistic target for the company which has failed so many times to deliver on promises. As CM always argues, the ‘cult’ following of Tesla is a dangerous vixen which can keep the ‘dream’ floating in orbit when reality is that “Nevada, we have a problem”.

The market can stay irrational longer than you can stay solvent. The 3% bounce in the shares reflects that blind optimism. Our study shows that even if it made margins similar to mainstream makers it is grossly overvalued.

Tesla X review


Catching an Uber to Sai Kung in HK and happened to be picked up in a Tesla X. The driver has had it for 9 months and said he’s had many problems. I asked him whether he’d buy another Tesla and he said no chance. He wants a Mercedes Benz. “Everything is better”.

His comments-

”handling – there is no confidence”

”autopilot – never use and friend has crashed using it.”

”quality is poor”

”noises in wheel hubs”

”centre panel is good”

”big windows good”

From the back seat the ride is thrashy but with so much torque suspension needs to be relatively firmer to stop squatting.

I get the coolness of this car. Indeed Elon Musk deserves praise for innovation and waking the sleeping auto giants to the luxury end of EVs. Sadly waking these giants that have production nailed means they hold all the aces, not Tesla. Tesla’s full year earnings aren’t far away where we’ll get a picture of whether the dream can stay alive.

As for this driver Tesla hasn’t won long term customer loyalty.

You can never be too connected


For all those that say men can’t multitask, this HK taxi driver can field 8 mobile devices and drive…what next a smart watch? Presume he is trading multiple markets. Either that or he’s sick of the stereotype that taxi drivers are usually the last to know!

Return to the nanny state – perhaps the ASB should consider envy not safety


Coming back to Australia often reminds me  that even when it is not a nanny state it is full of wowsers. This snippet is from Motor magazine. As one can see it highlights how some are compelled to whine over trivial things. Lexus, of all companies, is being bashed for encouraging speeding and the Advertising Standards Board has caved in (although I’m guessing it didn’t  take much to pressure them into the move) . Lexus was then bashed again even after removing the speedy bits.  I’m sure prospective Lexus buyers are waiting for footage of being stuck in a traffic jam

One would be pretty hard pressed to find a car company that doesn’t promote “performance” in its sportier range because by definition it’s kind of the point. Even Tesla for all of its supposed green credentials will brag incessantly about how it’s faster than any other road car to 100km/h, including exotica.

To say advertising “encourages speeding” just takes people for idiots. The moaners are surely aware that speed limit enforcement here is stricter than most. Speed cameras, highway patrols and red light cameras. Drivers here are constantly in fear of their lives. Autonomously driven cars will be a godsend as owners will not have to live in fear of being fined!

On top of that in order be able to buy a Lexus sports car like the LC500 probably requires $200,000. To afford such wheels presumably means that’s someone has half a brain. In traffic congested Tokyo Lamborghini and Ferrari are commonplace . The most impractical vehicle one can imagine but they still sell a tonne of them.

Perhaps we should look at the bright side that only the car company was only criticized for promoting speed rather than being hauled over the coals for not promoting the car in Saudi Arabia with a newly liberated female driver. We shouldn’t hold our breath. When Lexus advertise the LC500 again the next complaint will probably be that it focuses too  much on white males rather than diversity.

Alfa Romeo Giulia Quadrifoglio roadtest


I had a great opportunity to test drive the new Alfa Giulia Quadrifoglio (Q) over the past few days.  It is hard not to possess some sort of expectation before boarding. They say you aren’t a true petrol-head until you’ve owned an Alfa. To be honest Alfa hasn’t produced a rear drive monster for a very long time so the hurdle was pretty high on the basis of failure to introduce a Marcus Aurelius style gladiator to an arena filled with Germanic warriors would see thumbs pointed down.

The conclusion

Having tested the latest BMW M3, the one thing the Q  has over it is character. Sure it’s easy to dismiss the Germans for lacking a sense of humour as much as silencing Italians by tying their hands behind their backs. The fact is that the engine inside the M3 is a masterpiece. The handling is sublime but the exhaust note is synthesized through the speakers  which sort of disappoints like a Greenpeace activist would if forced to sail in a harpoon boat to stop Japanese whalers. The Q has 80hp more from a similar sized engine. It makes a naughty rasp as the tacho sweeps through 5,000rpm and has a tactile nature about it. Where the Q really wins out is ride. I’m sorry but for the mid-life crisis set, harsh ride is something that becomes annoying when you just wish to cruise on a highway late at night to get to the destination. The Q on soft suspension setting is more than acceptable. The BMW is just too jiggly. It is fidgeting. That’s ok doing a track day but when tired and cranky it is something you want to live without

The engine


A 505bhp 2.9 litre V6 twin turbo which strangely shares the same bore and stroke of the V8 Ferrari California. The twin turbos mysteriously sit between the V like the Ferrari too.  The engine has power anywhere in the rev range. There are 4 modes. Race which switches everything off, D for dynamic which is hyper, N for natural which is hyper (but less abrupt) and A for advanced efficiency (shuts down cylinders) which is great for pootling around town. The suspension can be set to soft on any of the modes meaning you can hoon in comfort.


Naturally it is brisk. The 8 speed box is responsive and in manual mode doesn’t try to second guess you. Perhaps 8th gear at the national speed limit in Japan (100km/h) is too tall given the engine is revving at about 1,650rpm (like a diesel). The engine happily pulls but it has a slight drone. Perhaps the car pleading to get it to at least 2000rpm where the NVH reaches a comfortable level. Not a fault of the car. When you need to overtake the response is instantaneous.


While never giving it the full beans on public roads the handling in the twisty stuff is telepathic and communicative. There is loads of feedback. Bumpy corners don’t upset it and you get the sense that if there were no speed cameras you could do twice the speed into those bends. The car also seems to have a neutral oversteer bias.


The vented 4 pots have oodles of stopping power. Fade free. Perhaps the only slight issue is when trying to modulate the brakes at a set of stop lights to smooth the ride for passengers you end up stopping with a small jolt.



Italians do design better than most. The interior is beautifully presented. Functional and easily hooked up to the iPhone although sometimes the interface is not as intuitive and the voice recognition technology requires work (I found out this isn’t the car but my default Siri setting that is the problem). The lady in the machine is Australian and even conversing with her in a native tongue trying to select an address gave me the directions to a coffee house in America. When she pronounced Japanese words in Her Aussie drawl it was hard not to burst out in laughter as to where she wanted me to go – for instance “Iidabashi” (pronounced E da bash E) was Oida besh E and deguchi (exit, pronounced deg uchi) was “day Gucci”. I didn’t try her in Japanese but I’m sure the problem would be rectified. This is a minor complaint. The phone hooks up quickly and phone numbers are all available for voice activation. There is a scratchpad which allows you to shift through dials. Seats are comfy and fully electric with lumber and heating. Steering wheel is also movable.



I’ll be honest. It isn’t the prettiest Alfa ever made but it looks purposeful. It sits low on its haunches and the carbon trim along the sills is tasteful not gaudy like some Korean airport  taxi that has crashed into an Autobacs accessory store.

In summary

This car has faults which gives it humanistic traits. it wouldn’t be an Alfa without. When on full lock the car makes a rather disconcerting shudder. However the faults it has are irrelevant when you assess what it is about the purpose. In a day and age where we have systems that help us stay inside lanes, slow us down, brake for us and remove the risks of human error, the Q wants to give the reins back to the driver if they choose.

It is also the feedback. While a Tesla night out drag it the Q will make one smile every time they sit inside. When you rev the engine at standstill the car body will twist. It is nice to know that there are carmakers who are unashamedly looking to petrolheads and ignoring the boredom of cars that breathe health and safety at every turn. Sure the Q has all the gubbins but you can switch it all off. Therein lies the issue.

At the Tokyo Motor Show I spoke to the Honda motorcycle marketing team. He asked what they could do to make their product better. My answer – “make it worse”. Astonished, he asked why? I replied that “riding a Honda is like sitting on a swing machine. It is smooth and efficient but there is no character. When I had my KTM Superduke 1290 it always let me know it was alive . Why not introduce software that makes the engine run rougher and edgier so the rider can engage with the machine and for the boring ‘otaku’ (geek) they can have their sewing machine!”  “so you’d pay to make it worse?” YES!!

While it is likely the M3/M4 will outsell the Q in sales volumes the reality is that the true drivers in their hearts would choose the Alfa. While Alfa’s of old had build quality issues, this felt tight as a drum. If Ferrari built a proper 4 door this would be it. It is by no means a poor Maserati either. It is a car long missing from the line up and for a return performance they’ve done an outstanding job. At $20,000 less than a BMW M3 the value proposition is also astounding. It is truly tempting!

Perhaps the sweetest function of all is in the settings where the car will tell you how economically you are driving. Whereas an electric car might coax you to drive to maximize battery life the Q encourages you to blow it off the other scale – in that sense it is a devil for environmentalists (even though fuel economy is vey decent) and that above all makes me want it more!

Tricking the auto-pilot 73% of the time


So much faith is put in the hands of computers nowadays but the idea of driverless cars is still fraught with danger.  Car & Driver reports “Researchers at the University of Washington have shown they can get computer vision systems to misidentify road signs using nothing more than stickers made on a home printer. UW computer-security researcher Yoshi Kohno described an attack algorithm that uses printed images stuck on road signs. These images confuse the cameras on which most self-driving vehicles rely. In one example, explained in a document uploaded to the open-source scientific-paper site arXiv last week, small stickers attached to a standard stop sign…using an attack disguised as graffiti, researchers were able to get computer vision systems to misclassify stop signs at a 73.3 percent rate, causing them to be interpreted as Speed Limit 45 signs..”

Sure systems will improve over time but we already have a plethora of people already putting too much “blind” faith in systems being fool proof as this video demonstrates