Thread by @ChrisBloomstran

@wintonARK Brett, Your additional thoughts on $TSLA’s insurance business are appreciated. That said, I respectfully don’t see where your comments clarify or dispel the points I made in response to the 2025 ARK model and price target for Tesla and its “insurance operation.” 1/

https://twitter.com/wintonARK/status/1373759146420400129

Please allow me to ask a few direct questions for clarification.

1. You confirm Tesla is not yet underwriting insurance & won’t be until 2023. Can you walk through your assumptions that get to your “bear case” $23B insurance revenues at a 40% EBIT margin, $9.2B by 2025? 2/

2. You project Tesla will underwrite at a 70% loss ratio. That’s actually not a better ratio than the US private passenger auto loss ratio over time. In fact, for the three years prior to 2020 the industry loss ratio was under 70%. 3/

The loss ratio is only one component of insurance underwriting. 2020’s was aberrantly low due to lower frequencies of accidents. How can Tesla operate at a 70% loss ratio and earn a 40% EBIT margin? Seems impossible to me. What will the expense and combined ratios be by 2025? 4/

Will any excess profitability earned by Tesla be regulated downward by state insurance commissions or competed away by an industry not lacking in capital? This is a crucial element seemingly lacking in your insurance model. 5/

3. In how many states to you project Tesla being admitted as an underwriter by 2023 to 2025? Will they underwrite in China and other countries by then, if at all? Will Tesla underwrite on any other vehicle brands? Insurers have more vehicles insured than policies-in-force. 6/

4. The ARK price target of $3,000 with a bear case of $1,500 and bull case of $4,000 presumes exactly 1B shares outstanding. At 12/31/2020 there were 960 million outstanding (905m the prior year). Your model presumes only an additional 40 million outstanding by 2025? How? 7/

There are 165m option and RSU shares already issued. If the stock price moves up to your range, presumably options will be exercised, correct? Will Tesla buy shares back despite requiring capital to grow? I don’t see how only 1 billion shares will be out by 2025. Model flaw? /8

5. Your model assumes no additional share issuance, while $12.6B was sold in 2020. Is the $19.4B in cash on the balance sheet at yearend sufficient to finance all capex and capacity increases? How will the company finance new factories and equipment? /9

Are the two operating plants and two under construction sufficient to manufacture your projected range of 5-10 million vehicles? Debt and preferred outstanding total $13.7. How much additional debt will the company incur by 2025? /10

6. How will $TSLA capitalize its yet nonexistent underwriting operation? Writing $23 billion in annual premiums requires at least $8B in statutory surplus. What will it cost to de novo seed underwriting in each state Tesla will sell auto insurance plus underwriting globally? /11

If Tesla will not issue new shares per your model, where will the capital come from to write $23B (ties Allstate as the 4th largest US auto underwriter, only behind State Farm ($41B), GEICO ($35B) & $PGR ($32) in depressed 2020 premium volume (due to COVID rebates/credits)? /12

7. You discuss Tesla’s advantage for, “Integrated telematics and data, inclusive of interior and exterior vehicle cameras, radar, and driver profiles.” Do you know what companies like $PGR are doing on this front already, with decades of actuarial underwriting experience? /13

What limitations do states like CA (where Tesla brokers an immaterial number of policies) have in place for use of technologies such as telematics and usage-based insurance? What are Toyota and GM doing now in already offering policies to customers? /14

8. My understanding is Markel no longer acts as an underwriting front for Tesla. Why? Filings in IL, TX and WA indicate subs of American Family will underwrite policies brokered by Tesla. Do you have commission projections for 2021 and 2022, before $TSLA begins underwriting? /15

9. You suggest zeroing out a projected value for insurance as immaterial to expected value if one doesn’t buy your projections. How are $23B in projected annual revenues and $9.2B in EBIT profit immaterial to a company that only did $31.5B in revenues and $2.1B EBIT in 2020? /16

I understand that the insurance revenue and profit forecast is immaterial in the context of a mind-boggling $3T projected value with a range from $1.5T to $4T. $3T in market cap would be ~12% of 2025 US GDP and 50% larger than $AAPL’s today. /17

Of course Apple does $300B in revenues TODAY, almost 10x Tesla's revenues. Your projections for insurance revenues and profits are material to any metric outside your extremely optimistic 5-year forecast for EV sales and commensurately gigantic valuation. /18

10. I appreciate your leaving solar & battery storage out of your model for now. I’d be curious how you view the capital intensity of investments here? Berkshire Hathaway is making enormous investments in wind, solar and energy transmission. No doubt alternatives are growing. /19

Coordination among FERC, state PUC’s and local bodies is complicated and generously yields only a decent regulated return. I see many bulls on Tesla making wild projections about the size of and profitability of Tesla’s investments and prospects here. I’d love your thoughts. /20

I look forward to your answers and any model clarifications. I hope you can see where a disbelief regarding a non-existent insurance underwriter growing to $23B and $9.2B in EBIT in less than five years perhaps may discredit ARK’s entire model and assumptions. /21

As immaterial as those numbers are to your forecast, they are material to Tesla as exists today, material to its capital needs prospectively, and extremely material in the highly competitive world of auto insurance. /22

Further, if there are 147 million option shares outstanding and 18.8 million RSUs, modeling an increase of only 40 million shares over the next 4.75 years to an even 1 billion seems shaky at best if not fatally flawed. /23

At your $3,000 price target, the uncounted 126 million shares (total if all exercised would be 1.126B outstanding) ignores a not insignificant $378B in market value, 12 times current sales and $11B MORE than your BULL CASE 2025 EV revenues of $367B on 10 million units sold. /24

Unless Tesla is done issuing option shares and RSUs or you project a massive share repurchase program as the company spends capital to scale up, the failure to properly model share count further weakens the credibility of the ARK model and price projection. /25

One final note. I know you are leaning on a Monte Carlo simulation to piece together your model and posit that, “the 25th percentile outcome is not the downside for each of these inputs. The bear case of 3 coinflips is not 3 tails, but 1 heads and 2 tails.” /26

Where I come from, risk is permanent loss of capital. Bear cases in my world are harsher than positing a price gain from $655 share to $1,500 over 4.75 years. Most rational investors don’t bake in a 19% annual investment gain over a 4.75 year horizon and call it a bear case. /27

Given $3 million and three coin flips at $1 million per flip, a properly estimated bear case is a $3 million loss, Monte Carlo be damned. With Tesla’s ability to raise cash by selling shares to ebullient or passive investors, the risk of bankruptcy is now far removed. /28

The bear case, however, can’t be a $1,500 stock price and 19% annual gain from Friday's close. Bad things need be accounted for in bear cases. The ARK bear case seems absurdly optimistic. Your reply to the above questions is welcomed. Thanks again for responding initially. END