You can install our site as a web app on your iOS device by utilizing the Add to Home Screen feature in Safari. Please see this thread for more details on this.
Note: This feature may not be available in some browsers.
@View attachment 410901 View attachment 410902
In an effort to make a simple, but compelling forecast of plug-in EV market share, I have boiled things down to a random walk on a logit scale. May sound complex, but it actually makes very few assumptions. I examine market share from 2012 to 2018, computing the annual increase in logit market share (log(S/(1-S)). So I've got now 6 data points. I compute the mean and sample standard deviations. I make the assumption that future logit differences have the same distribution as was operative in the past. Thus, the future is modeled as a random walk with a certain mean and standard deviation. I can than predict the mean and variance (inclusive of parametric uncertainty) of the logit random walk. From this I can convert back into the market share scale, tracing out the mean path and a 90% predictive envelope.
It is the predictive envelop that I have not really explored before, but it is critical for understanding just how much uncertainty there is in a forecast that is properly conditioned on historical observation. If another modeler were to put forward a forecast that is substantially outside of the envelope, I would know that they are inserting information into there analysis that is not consistent with what has actually been observed since 2012. This does not mean that such a forecast is "wrong", but merely that it represents a substantial departure from historical trends. This departure may well originate in the modeler's own imagination about what make speed up or slow down the gain of market share in the future. For example, vehicle autonomy could alter the environment for EVs and ICE in ways that were not present in 2012 to 2018. Then again, Tesla has led with AP which may in fact give their cars a competitive advantage already witnessed in recent past. So these matters are largely a judgment call. I view my job as a modeler as to reveal what the data is telling us and to try not to assert my own opinions as an overlay to the data. This is the virtue of a truly statistical model over a judgmental model.
You can see the fruit of this simple analysis in the charts above. I have presented this with the y-axis as either on the nominal (market share) scale or the log market share scale. In the former, we see a typical logistic curve surrounded by the predictive envelope. Note that this envelope is widest around the year 2028. Yes, this is actually where we have the greatest uncertainty! Notice also that the envelope passes through the 50% mark between the years 2026 and 2030. Anything that might substantially accelerate or delay EV market dominance had better happen by 2025, otherwise it is just too late to make much of a difference. So if we imaging that autonomy will speed it up or "lack of public charging" (Come on, BNEF, you're better than that!) will slow it down, those things need to come into play within the next 5 years or it just won't impact he timing of market dominance
But some will look at that chart and think, "How can those tiny little historical observations blow up into such a big effect? I just can't believe that." This of course is the problem people have with intuiting exponential (or logistic) growth. So the whole forecast will strike them as a fanciful extrapolation. This is why I also present the exact same data with a log scale for market share. What is striking in the log scale is that one can see how the historical growth is strongly linear (or rather, log-linear). The pattern has been remarkably linear, and this is precisely why a random wake with drift is a compelling model. One is invited to question what exactly could take log market share off this strongly linear path. Indeed, one must see this clearly to understand why only very strong forces would really be able to knock it off course. For example, the oil crash of 2014-16 hardly makes a dint in the historical trend. To be sure it is there, but it is such a minor effect that one must look very closely for it. Of course, we know in the long run market share cannot exceed 100%, so ultimately the line must level out asymptotically. But notice that the bend does not really make much of difference until after EVs have dominated the market. Up through about 2025 EV growth will not be distinguishable from exponential growth.
In reality all forecasting is extrapolation. But if I must extrapolate, I prefer to extrapolate from data more so than from opinion. The trend is clear while the window to substantially alter the trajectory is narrow. When the data is bending to a different trajectory, I will gladly change my opinion. But for now the data are not showing signs of slowing, if anything the path is mildly speeding up.
Now let me make some predictions. 2019 PEV share comes in between 2.7% and 3.7%. Believable? How about 2020 between 3.8% and 6.1%, or 2022 between 7.4% and 14.8%. These may seem fairly wide, but not so wide as to be without consequence. Consider that BNEF is predicting on 10M EVs sold in 2025 or share of 10%. My model puts 2025 between 19% and 43%. So BNEF is already 2 standard deviations below my lower bound, which is about 2 standard deviations my mean of 29%. So BNEF is seriously bending the curve down in ways to which historical data does not bear witness.
Or let's back test this. In April, 2018, BNEF forecast 2018 to come in at 1.56M or about 1.67% share. The actual was 2.018M or 2.12% share. Let's what my method would have predicted using just 2012 thru 2015 data (a sample size of just 3). My 2015 forecast of 2018 would have centered on 2.2% with a 90% predictive interval from 1.3% to 3.5%. Yeah, a lot of uncertainty, but nailed it. My 2016 forecast of 2018 centered on 1.8% ranging from 1.3% to 2.5%. This was a little more pessimistic, but less uncertain. Then the 2017 forecast of 2018 centered around 2.0% ranging from 1.7% to 2.4%. So the predictive envelope nicely closed in on the actual. Meanwhile, BNEF low forecast was ruled out of the predictive interval by the time it was made using 2017 as the last historical datum. Presumably, BNEF's forecast had that benefit of highly granular data, their proprietary data, and a multi-industry team of analysts. Surely with all that going for them, they should have been able to produce a forecast with much less uncertainty, but in fact a sample size of 5 historical observations could have alerted them to the possibility that their prediction was high improbably.
So I am not at all saying that my very simple model is the best. That's not the point. The point is we need simple challenger models that don't assume much but capture the uncertainty of the historical record to tell us when our fancy forecasts are bending to improbable conclusions, overburdened with too much complexity, to much granularity, and to much opinion. If any modeling group out there wants to avoid embarrassing themselves with precious predictions, they would do well to stay with a predictive envelope such as I have constructed. This is what we call a challenger model. And to the rest of us, if we want to avoid being lured in by "credible" forecasts from well-funded organizations, we do well to have a few simple challenger models of our own.
I want to apply one more variation to the China forecasts. In scenario generation, it is a common practice to use one model or forecast for the mean structure and then build out an error structure on than mean forecast. What this comes down to is that I want to start with Dr Holland's prediction of 7.5% in 2019 and 50% in 2025. I want to take that as a given, as a hypothesis. I literally connect the dots on the logit scale to interpolate from these two points. This is the assumed mean forecast. Then I assume that annual standard deviation (on the logit scale) is 20%, which I derive empirically from the historical data points. So I assume an error structure about the mean that is a random walk. So in the chart below, I provide this predictive envelop along with Dr Holland's (DMH) interpolated scenario.
View attachment 411790
View attachment 411789
So we see that this scenario envelop is substantially more narrow than the predictive envelope I built based solely on historical observations (JHM). This is because the JHM envelop includes the uncertainty about the mean growth rate. Indeed on the logit scale the mean is 48.39% with a standard error of 10.09% which is in addition to the volatility of 20.19%. So that is a fair amount of noise to signal. That is good forecasting. But in scenario generation we are allowed to create scenarios to that do no represent all the uncertainty. So one way to think about the DMH scenarios is that under the hypothesis that Dr Maximilian Holland has nailed the mean path, random paths are likely to remain within the envelop 90% of the time. So think of it more like a hypothesis test with the mean path as the null hypothesis.
So this does demonstrate the value of both kinds of modeling approaches. Dr Holland has done the difficult work of assessing the feasibility of his prediction. My predictive envelope completely includes the Holland's scenario envelope. So that suggest that his approach is entirely consistent with the historical observations and of straightforward interpretation of that trend. This is what good detailed modeling should do. Had Holland's scenario drifted outside my envelope, I would have become concerned that he is projecting things that have little basis in recent history. For example, BNEF is claiming that EV penetration (globally) will only be about 30% by 2030. This is outside of my predictive envelope. Not only is it pessimistic about EV adoption, it is improbable too. Had they predicted 50% by 2030 instead, then I'd say that, while is pessimistic, I do not yet have sufficient data to rule it out as improbably pessimistic.
Simply put, I can look at Holland's prediction as say, "yep, that looks like where the data is headed." But looking at BNEF, I shake my head think, "No, this opinion is at odds with the historical trend. The observed data are clearly not headed that far in that direction." Moreover, Hollands scenario is incompatible with BNEF. There is no random path that can remain probabilistically close to both forecasts. I.E., one of them is substantially wrong about the future.
Notice also that my JHM-Mid path is within the envelope of DMH. So my mean structure is not at this point in time a statistically significant departure from Holland's mean structure. My guess is that the way Holland is approaching this is first to observe where the trend is headed, and then look a detailed data to see if there is some material reason to shift the forecast up or down from that basic trend. So long as the battery, auto and policy makers are making commitments that come close to realizing the projected growth rate, there is little reason to bend that projection one way or the other. Intellectually this is anchoring the analysis on recent trends, while being ware of issue that may force trends to change. I suspect that BNEF is more skeptical of historical trend and seeks to anchor a forecast on something else. But when you look more carefully there are often compelling forces that have shaped historic trends, and these forces are dismissed at intellectual peril. For example, consider all the investments and commitments that EV producer and EV consumers have made just to bring about recent history. These are long term commitments. These are things like building a Gigafactory or spending $1000 just to add a charger to your garage. Those who make these commitments are not stupid, and they reasonably expect them to pay off in the long run. So to suggest that the growth trajectory will slow down in some way not yet manifesting in recent history is to suggest that all these investments were ill-considered and will not payoff as expected. The auto industry largely works on a 6-year product development cycle. Of course, not all new products are going to be commercial successes, but the time scale is such there ought to be substantial momentum to the whole industry. The industry can't radically speed up relative to recent history, and it can afford to substantially slow down either. If you take this kind of investment momentum seriously, then you have to accept that trends will largely persist. They can bend a little, they can have short-term hiccups, but they have serious momentum. It is like watching a train try to slow down or accelerate. There is so much momentum at play. There is only so much braking or accelerating that can happen over a certain interval of time. You can predict the train wreck because you know the train can't slow down fast enough to avoid collision. This is why trends are so important. It is the same sort of momentum that makes it so hard for the world to actually reduce carbon emissions growth. It takes huge trillion dollar investments just to halt demand growth for fossil fuels. And the nature of investments is that they do grow exponentially where the opportunity for high growth rates exist. This is exactly what locks EVs into a logistic growth curve. The level of investment grows with the size of the market, fueling exponential growth slowing down only as the market becomes saturated.
Since I digress here, let me point out one more thing. Policy accommodation easily follows the expansion of an industry. Some energy modelers are inclined to base their energy forecast primarily on policy commitments. But those policy commitments are easily a lagging indicator rather than a leading indicator. For example, EV incentives are likely to decline as EVs become more competitive in the market. Does that mean EV sales will decline? No so fast. If they are declining because it is determined that the are not longer needed, because they have outlived their usefulness to society, then this is policy lagging success. While there can be short-term policy effects like pulling forward sales before a subsidy expires, there need not be any durable impact. For another example, some countries are looking to ban new ICE sales by a certain date. This too is lagging indicator. Few politicians would venture such a policy unless EVs already looked thoroughly capable of replacing all ICE sales by that certain date. For example, no one is legislating that by a certain date all vehicles sold must be hydrogen fuel cell powered. That would be crazy talk because the fuel cell car industry no where looks capable of fulfilling such a mandate. So policy makers are in fact looking at what the EV industry has been doing in recent history and extrapolate from there what may be politically feasible policy. This is very much politicians lagging industry.
So with China, there is all this concern that political support for NEVs might wane. But why should it. To the extent that EVs become economically competitive apart from subsidies and mandates, we can accept that the government may pivot to other areas where incentives may have a bigger impact. This is just the training wheels coming off. But as China recognizes that they sit on an enormous opportunity to export EVs to other countries, we have got to be expecting that the government will solidly push in that direction. So yes, policy matters, but policy is just part of the bigger process. It will participate in market opportunities as they unfold. Policy is largely an endogenous variable. So policy is as much part of the momentum as it is a catalyst for change. So I am not at all worried that policy support will wane and cause decline in EV adoptions. If anything I believe that policy will become more assertive as EVs advance. Why wouldn't China want to be the world's largest and most important exporter of EVs and batteries?
So momentum matters. And China looks to be EV dominant by 2025.
Thank you for posting this thorough analysis by @jhm. As usual, he is clear and concise (and probably more reliable and realistic than BNEF).For comparison:
Tesla is so far ahead that it will be difficult for any of the majors to catch up. Look at the new Audi EV... it's just pathetic. Half hearted offerings from other majors have met with lackluster acceptance. Dealers don't want to sell them and their specs are weak.If Bloomberg is right then IMO, Tesla has awoken the sleeping giants and although they have been incredibly slow to respond, late 2020 could see a scenario where the buyer of an EV has legitimate options without long waiting lines from companies who have economies of scale and vast manufacturing and distribution channels that have been serving customers for 100+ years. This is something Tesla has never had to contend with.
I suspect the coming EV war will be dirty and there is no way Tesla will emerge from this fight without being knocked about, bogged down and desperate for cash.
Tesla is so far ahead that it will be difficult for any of the majors to catch up. Look at the new Audi EV... it's just pathetic. Half hearted offerings from other majors have met with lackluster acceptance. Dealers don't want to sell them and their specs are weak.
I do agree that the EV war will be dirty. It already is. But it's not between different EVs, it's between EVs and ICE. Why do you think Tesla is constantly being attacked with FUD?
If Bloomberg is right then IMO, Tesla has awoken the sleeping giants and although they have been incredibly slow to respond, late 2020 could see a scenario where the buyer of an EV has legitimate options without long waiting lines from companies who have economies of scale and vast manufacturing and distribution channels that have been serving customers for 100+ years. This is something Tesla has never had to contend with.
I suspect the coming EV war will be dirty and there is no way Tesla will emerge from this fight without being knocked about, bogged down and desperate for cash.
Skyrocketing Tesla Sales Force Mercedes Dealer In Norway To Face A Kodak Moment | CleanTechnica
It is perhaps a good time to review the graph here (on the bottom) by Nicholas Felton. Norway is far ahead in the EV adoption curve, but we expect other countries will be in the same place within 5 or so years.
This should serve as a warning to companies like Toyota that take a “wait and see” attitude toward EVs. As Tony Seba describes more clearly than anyone else I know, although a disruption starts slowly, when things hit a tipping point, demand shifts quickly and those unprepared for the change must suffer the effects of the disruption.
although they have been incredibly slow to respond, late 2020 could see a scenario where the buyer of an EV has legitimate options without long waiting lines from companies who have economies of scale and vast manufacturing and distribution channels that have been serving customers for 100+ years..
Fact remains that Tesla will have serious competition and cash problems in the near future and I would not discount MB, BMW and VW's commitment to EVs as quickly as I would GM and Ford. You can believe AlixPartner or not but Tesla will not sell all these cars. - "pure electric models will account for 20 percent of U.S. sales by 2030 while reaching 30 percent in Europe and 35 percent in China..."Since I don't understand this language, I put your post in Google translate and this is what I got:
Blah blah blah... bluh bluh bluh... blah blah blah... bluh bluh bluh. Na na boo boo
Fact is that your "facts" are just made up speculation without any basis in actual fact.Fact remains that Tesla will have serious competition and cash problems in the near future and I would not discount MB, BMW and VW's commitment to EVs as quickly as I would GM and Ford. You can believe AlixPartner or not but Tesla will not sell all these cars. - "pure electric models will account for 20 percent of U.S. sales by 2030 while reaching 30 percent in Europe and 35 percent in China..."
IMO, if during that first time period, there is a jump to light speed in a technological improvement that lowers the price, the demand will be extraordinary for EVs.
Fact is that your "facts" are just made up speculation without any basis in actual fact.
None of the traditional automakers have been able to produce anything that comes close to the 2012 Model S. They are 5-10 years behind.
The only expertise they have is in making IC engines and transmissions. Everything else is farmed out. Unfortunately, ICE is obsolete and useless when making EVs so they have to start over. They have been reluctant to make the investment, hoping that EVs would just go away. They are now being forced to make EVs but their starting line is many years behind Tesla. I do hope they get their act together and start making some compelling EVs but I think it will take at least another five years. By then, Tesla will have moved forward another five years.
Also, not sure why Tesla would have cash problems when they make a profit on every car they sell. If they were in the position of VW group, BMW, Mercedes, etc. where they lose money on their pathetically small production, it would be a problem.
Lots of promises and speculation about " just wait five years" then we'll really compete.Y
You need a dose of facts. First: Tesla's real cash position.
Tesla is burning cash. Tesla sold new debt in large part to pay off old debt and the net proceeds of the $2.7 billion raised this month will work out to be only ~$0.8 billion assuming the past debt is paid and Tesla’s working capital balance is put to zero (read the prospectus).
Furthermore, Tesla has been underspending massively in its business. Tesla's publicly announced that its capital expenditure budget for CY 2019 would be $2.0 to $2.5 billion ($680MM per Q), however, Tesla only spent $280MM in Q12019 and $325MM in Q42018. That is a deficit of $750MM. Tesla also spent $27MM less on R&D Q12019 compared to Q12018.
IMO, Tesla decided to forgo spending to save on cash.
Second other car companies commitment to EVs.
VW has announced that it plans to spend $10 billion on EVs by 2023 and have publically upped VW’s global EV sales projections from 15 million vehicles to 22 million over the next decade. VW has also stated that it will spend $800 million to expand its factory in Chattanooga, Tennessee, to handle that crossover and another all-electric model.
Mercedes has stated that they will be launching 10 pure battery-electric vehicles by the end of 2022. Unknown how much $$ they are forecasting to spend outside of China.
As of two months ago, Ford has a new CEO and he is on record stating that Ford plans to spend $11 billion on electrification technology by 2022 to develop 40 new all-electric and hybrid models.
GM basically is nowhere except Bolt, Kia has the Niro, Hyundai has the Kona and IMO, the Bolt is a better car. Honda has cold feet, Toyota’s first all-electric cars will go to China and it is unclear when if ever the USA will see one. BMW is spending an extra $1B on EV R&D this year and has promised 12 EVs by 2025.
To sum up, Tesla changed the world, has great cars but is underspending and burning cash rapidly every quarter. If Tesla had the cash VW did, it would not be a close race but they don't and IMO, the once silent majority is now picking up speed.
Lots of promises and speculation about " just wait five years" then we'll really compete.
Reality...
VW to Reshuffle $56 Billion Battery Push as Samsung Deal at Risk
BMW Car Unit Posts First Loss in a Decade
By
B- Are you a robot?
Why Volkswagen's Profit Dropped in 2018 -- The Motley Fool
Ford profits down more than 50 percent in 2018
Donald Trump Over-Cheered an Auto Industry That Has Already PeakedDid you ever take a look at these company's financials? Ford, VW, MB, BMW, GM - They are all profitable companies with strong cash reserves. Even F made $1.1b in net income Q12019, Daimler $2.1b, BMW $0.5B, VW $3b, and GM $2b.