View Single Post
      03-03-2017, 08:37 PM   #1
M.FriedmanFast
Private First Class
52
Rep
144
Posts

Drives: 2013 F10 M5
Join Date: Feb 2017
Location: Bay Area

iTrader: (1)

Demystifying the M-TAX - Carbon Fiber

Hello All,
I am a new poster to this forum yet a long time purveyor of its valuable information. In the coming weeks, I will post pictures of my stable. The purpose of this post however is to demystify and issue a rebuttal to the M-tax as it relates to the performance modification market. This subject has been discussed ad naseum, however I am unsure whether or not an empirical and economics-based approach has been employed to analyze what is at best opportunistic capitalism and more likely, egregious price gouging (also opportunistic) due to (1), the relative lack of liquidity in the market as it relates to private investment (of any scale at least) and (2) the irrationality we all demonstrate as market participants who want nothing more than to customize and improve the performance of our vehicles. To reveal biases, I am a strict adherent to neoclassical economic principles and therefore support free market enterprise to the fullest; what I loathe however, intellectually and practically, is market inefficiency and arbitrage that need not exist. Suffice it to say that this market is of the most inefficient in any sector, based chiefly on the two above premises. To fully deconstruct this issue would take far more time than I have to dedicate to the subject, and as such, the analysis utilizes the manufacturing and marketing of carbon fiber (“CF”) aero pieces, specifically a front lip spoiler, as a case study that is microcosmic of nearly every segment of the performance modification market with the lone exception of ECU tuning

The market for CF is perfect example of bifurcation. There are the premium brands and the value offerings, of which there are myriad, the majority of which are manufactured in China or other lower-cost jurisdictions. The price delta on average between the two market tranches for CF front lip spoilers is nearly 300%. Said delta is justified by the premium brands as being attributed to intensive R&D and ahem, engineering. The overarching hypothesis of this analysis a premium-equivalent offering can be designed and manufactured and marketed at a fraction of the cost whilst still generating substantial economic rents for the investor.

The first M-tax myth then (as it relates to this case study):

1. The design process is rudimentary and therefore is neither human capital nor necessarily monetary capital intensive

And similarly and more importantly here:

2. CF aero pieces present no aerodynamic benefit; certainly not those supported by proper computational fluid dynamics (“CFD”) modeling; if anything, they impede performance within certain air velocity regimes (don’t get me started on so-called diffusers – and yes, I have them on all of my cars for aesthetics only – examine the actual designs, if anything, they decrease flow velocity thereby creating an increase in floor pressure, Bernoulli anyone?)
Specific to myth #1: a second year ME or Aerospace undergraduate student would have amassed sufficient technical aptitude to design any CF spoiler on the market. Reverse engineering was rendered all but laymen due to the advent of CMMs (coordinate measuring machines) and more recently, 3D scanners. With either technology coupled with the associated software, one can, in a matter of minutes create a dimensionally accurate sold model for importation into Solid Works or the like. In this case, one would need to scan the front bumper of a vehicle. Once the solid model has been created, the “engineer” creates a bespoke design based on varying criteria (ease of installation, fitment, cost to manufacture etc.). The bespoke design process is challenging inasmuch as creativity is required, artistic creativity precisely; 95% Picasso, 5% Isaac Newton sounds about right.

Specific to myth #3: aerodynamic optimization is empirically achieved through a combination of wind tunnel testing with the appropriate telemetry and CFD. The monetary capital intensity of both such processes is significant, as is the human capital intensity. The modern platforms, be it ANSYS, Adina, Aerologic or the like, require technical expertise typically reserved for PhD Physicists or PhD MEs; sure, some hack can establish erroneous boundary conditions and run moment-simulation and claim that they performed CFD design optimization, but trust that no company of note is entrusting the commercialization of its products to that Rhode Scholar.

Because I actually like and procure parts from the several premium brands in question, the following anecdote is presented with anonymous protections. Somewhat recently, I interacted with a so-called head of product development for a certain premium brand and we embarked on a discussion regarding the complexities of designing aero components. Claims were made that defied the most basic precept of logic concerning design lead time, mold costing and wait for it, aerodynamic optimization, specifically drag reduction. Suffice it to say that the gentleman quickly found himself over his skis after fumbling 2 or 3 basic questions regarding drag reduction and the conversation abruptly ended. As he walked away, I quipped that I would love to see the CFD model itself based on the fascinating claims submitted. Crickets…

Ok then. If the design process is relatively simple and is not complicated by the vagaries of aerodynamic optimization, why do the premium brands command such inflated prices? Surely it must be the marginal cost to produce the components, namely the CF itself. And the third M-tax myth:

3. The marginal cost to produce even the highest-quality CF is not commensurate to average premium prices examining peer group gross profit margins; it ain’t that expensive

CF front lips average 500 square inches. Considering the predominant curvature (this could be another myth, designs are not that differentiated) and utilizing only pre-preg CF from world class manufacturers with armies of Six-Sigma black belts, I garnered quotes ranging from $0.45 – $0.85 per square inch based on a 5-year production volume of 400 – 500 units. I denominated the below analysis in inches for ease of scaling the model (I know, CF typically sells by the yard). So then the effective range on CF marginally is $225 to $425 per unit based on the above demand forecast. As for the mold, I garnered quotes ranging from $5,000 to $7,500. Said molds based on the demand forecast would have at least a 5 year life expectancy, most likely twice that. It should be stated here that the below economic analysis contemplates the max values in all incorporated ranges, this to preempt any accusation of cherry picking the lowest values. So, to baseline, on marginal costs, we are at an artificially high cost basis of approximately $450 per unit. Ok then, given the average premium brand price of approximately $2,000, it must be the excessively burdensome overhead (and I use this term lightly as to my knowledge, none of the premium brands are vertically integrated, which is to say producing their own CF, and if they are, well, the greater fool’s theory applies).

I employed a basic manufacturing cost model where overhead is absorbed on a unitary basis; in other words, a fraction of the total overhead costs are allocated to each unit sold based on a budgeted value per se. In this example, I assumed that the company, here a new market entrant seeking to disrupt the premium brand’s market share, generates 1,500 unit sales per annum. To render the most conservative analysis, I assumed that a one-time capital investment (“CAPEX”) was required to produce this spoiler, that existing and sunk costs could not be deployed (a dubious assumption, but stay with me). The investment consists of a Creaform Handyman 3D scanner, a Solid Works License, a laptop and of course the CF mold discussed above. To make it more realistic, I assumed that this CAPEX would yield 5 new products, 4 in addition to our mythological spoiler. I also accounted for a fully-loaded design facility in Northern California (talk about negative arbitrage, but again, conservatism) Based on these assumptions; each spoiler would be allocated a $135 overhead charge. Ok, considering the marginal cost to produce each component plus a healthy overhead allocation, cost per unit now total approx. $515. There must be something else, as based on this cost basis, the premium brand gross profit margin would be around 75% (that’s actually not as absurd as you would think).

Of course, Sales and Marketing. I accounted for an incremental full time employee whose sole existence is to generate 400 unit sales over the next 5 years. This adds an incremental $85 per unit on a fully absorbed basis (which is another fallacious assumption as the asset could be leveraged to do other tasks, such as cleaning toilets or wiping the drool off the dog's face, or the like).

Finally, pricing. Recall, I am demonstrating the economic feasibility of producing a component of equivalent if not superior quality for a fraction of the price. Arbitrarily, I assumed a 50% discount to the premium incumbents’ price, namely $1,000. For the purpose of net present value and IRR, I’m assuming a 100% equity financed business, hence the opportunity cost of capital, or hurdle rate for the investment is set to a meteoric 15%. The takeaways are as follows – the new market entrant would:

1. Sell the device for $993
2. Generate a 43% gross profit margin
3. Generate a 35% operating income margin
4. Generate $347 in pre-tax profit per unit
5. Generate $15K in NPV or a 45% internal rate of return (IRR is far more meaningful here given the exponential scaling of NPV on greater units sold, ergo more than a paltry 400 spoilers over a five year period)

Assuming cost parity, the premium brands are commanding a 72% gross margin, or a 28 point delta to our imaginary company. Surely they will argue that I have underestimated their cost structure, or perhaps even more incredulously, I am somehow uninformed. Let’s unpack and preempt this rebuttal: (1) I performed this analysis in 45 minutes between investor meetings – rudimentary stuff; (2) I acquired molding and production quotes over a 24 hour period; (3) I thought about citing my resume and current occupation here, but will refrain from doing so, generally considered poor form. In short, if you cannot structure a cost basis equivalent to what is modeled, then the University of Chicago is awaits you, or frankly any B-school of note 
In conclusion, this degree of arbitrage can only exist in a very inefficient and here, illiquid market.

Demystifying the so-called barriers to entry was a 1-day task (cycle time). The premium brands command the prices they do because we (myself included) continually buy their goods. Some of us do so because our marginal propensity to consume is very high (disposable income etc.). However I posit that the majority of their brand loyalty is a function of the façade of technical difficulty, trade secrets or any other vehicle offered to justify absurd pricing schemes. Perhaps at some point in the future when I am less occupied, I will speculatively invest here and improve the efficiency of the market (anyone can actually). Until then, I’ll wait for my new premium brand diffuser to arrive!
Attached Images
 
__________________
2013 F10 M5 Singapore Grey | BPM Stage-II Tune | SSP Spec-R clutch-pack| RPI DPs | RPI GTM | RK Intake | Custom Aquamist WMI system | KW V3 coilovers | Dinan Sway Bars | RW Diffsuer | RW Front Spoiler | IND Trim | HRE s101s | Passport 9500ci | AWRON digital gauge cluster
AND: e92 M3, e70 x5, 997.1 GT3, F15 x5M, e34 M5, 80-series LC
Appreciate 2
mister2d155.50
M5 London307.00