The expected value from a BioNTech and Matinas deal is off-the-charts bananas

A licensing agreement between BioNTech and Matinas could generate an absolutely epic amount of value. How much value? Well, unless you’re an omnipotent being, then figuring out the answer is impossible. There are too many variables. Especially since BioNTech supposedly wants to use Matinas BioPharma’s LNC technology for the entirety of their mRNA platform.

But we’ll take a stab at it anyway.

For science.

Here is the basic formula for expected value when it comes to betting:

(Payout * ProbabilityOfWinning) – (CostOfBetting * ProbabilityOfLosing)

Let’s say you were to bet $1 to win or lose $1 on a fair coin flip. It would look like this:

($1 * 50%) – ($1 * 50%)

($1) – ($1) = 0

If you made 20 bets, you’d probably end up plus or minus a few dollars, but over time that value will gravitate towards zero. Zero is the expected value.

This is how casinos make money. Every game in the building has negative expected value. Unless you’re playing blackjack and counting cards. This would give you an edge of about 2%. (Which is why most casinos don’t allow this.)

If a bet has negative expected value, you should never make it. (Unless you’re just at the casino having fun with your friends. That’s different because the expected value you lose from the bet is balanced out with the value you get from the entertainment.)

If a bet has positive expected value, you should almost always take it, assuming the cost of the bet doesn’t exceed more than 1-5% of your total funds. In gambling, this is called bankroll management. In investing, it’s called diversification. It’s pretty much the same thing

It’s best to avoid betting too much on one outcome. Even if you’re favored to win. This is because in life things can, and do, go catastrophically wrong. Or maybe you’re tired and you miscalculate. Or sometimes the game is rigged. Regardless, the takeaway is this: If a bet has positive EV, and the cost of the bet is tiny compared to your total funds available, you should almost always take that that bet. Even if it feels uncomfortable. Otherwise, you shouldn’t be betting.

Let’s look at Matinas and BioNTech

This first calculation will make some assumptions. First, that formulating an oral, shelf-stable, and potentially less toxic Covid-19 vaccine will capture 85% of the market. It’s probably higher but we’ll be conservative here.

For our first example, we’ll say the payout is $54.4 billion. $64 billion is the estimated 2022 Covid-19 vaccine market but $9.6 billion gets syphoned off by competitors due to existing deals, marketing, and people being afraid of new technology.

We’ll say that the chance of successfully formulating an LNC mRNA vaccine is 10%. Given that Pfizer/BioNTech Covid vaccine has already been proven to work, and that LNC tech was recently proven to work, I think most people would agree that 10% is a very low chance of success. This isn’t a new drug they’re building we’re building from scratch, it’s a superior formulation. The actual chance of success is probably closer to 100% but whatever. We’ll use 10% anyway so you can see how ridiculously stingy BioNTech is reportedly being in their negotiations with Matinas.

We’ll say the upfront payment on a licensing deal is $100M.

($54.4B * 10%) – ($100M * 90%)

So now we get $5.44B - $90M, which is $5.35B

Most licensing deals include milestone payments and royalties. Let’s say that amounts to a $500M.

This means the expected value for the bet is $4.85B

If this was a casino game it would mean that for every $1 you bet, you can expect to win $48.5.

Which is insane. The casino would be bankrupt. Again, these are conservative numbers. Although the market for Covid-19 vaccines is predicted to shrink. Let’s say it shrinks by 90% and the chance of a success falls to 3%, but the LNC mRNA vaccine captures the whole market.

(6.4B * 3%) – ($100M * 97%)

($192M) – ($97M)

This brings us to $95M. With an upfront licensing payment of $100M and then milestone payments of $500M we’re now at -$505M.

Does this mean BioNTech should pass if that’s what they think the numbers are?

No. Because the Payout is a yearly payment whereas the CostOfBetting is a one-time expense. This is where investing calculations differ from betting calculations.

Year 1: -$505M expected value

Year 2: -$313M expected value

Year 3: -$111M expected value

Year 4: $81M expected value

Year 5: $273 expected value

The longer sales go the more money you make. Drug patents in the US are technically for 20 years (unless you get a special designation), but it depends how long it takes you to test your drug and get it approved. For Covid-19, this was shorter than usual due to the pandemic emergency.

So, even at a laughably low chance of success, and a greatly reduced Covid-19 vaccine market, BioNTech should still pull the trigger on a deal with Matinas because in the long run the expected value is still positive.

Although if BioNTech can find a better drug delivery system with higher expected value, they should go with that one. But since there isn’t anything on the market that I’m aware of that’s even close to LNC, I doubt this is the case.

If anything, the two companies are probably arguing over how much value LNC brings to the equation. Matinas would argue that it’s high, and BioNTech would argue that it’s low. Which is expected since it’s in both company’s financial interest to do so. Which again, is dumb, because this answer is unknowable. It can only be seen after the fact when an LNC vaccine starts taking market share from competitors. If Moderna has half the market for Covid-19 vaccines, and then Pfizer/BioNTech introduce an oral vaccine that’s shelf stable and more effective, allowing them to capture 100% of the market, then LNC’s contribution is 50%. Hopefully the parties negotiating this deal understand that.

But let’s get back to the total expected value on this deal.

Here is the borderline worst-case scenario:

The conversative calculation was just for Covid-19 sales. Don’t forget, BioNTech supposedly wants a license for the entirety of mRNA. Which is where negotiations go into fantasy land. It’s not possible to estimate the expected value of such a deal because there are trillions of variables.

BioNTech’s pipeline is enormous. They have 24 mRNA products listed on their website. Calculating the exact Payout value for this is only possible if you live in the clouds above Mount Olympus.

But again, let’s just be ridiculously conservative and say that each Payout (the size of the market for each drug) in the pipeline is only $1 billion. We’ll even shrink the Covid-19 vaccine sales down to a billion. And we’ll drop the chance of success for every product to 3%.

This brings the Payout to $24 billion.

(24B * 3%) – ($100M * 97%)

(720M) – ($97M)

This brings us to $623M - $500M in milestone payments for an expected value of $123M. But that’s just in year one.

Year 2: $843M

Year 3: $1.56B

Year 4: $2.28B

Year 5: $3B

This is the disaster scenario. One drug is successful and approved for sale. It has yearly sales that peak at $1 billion and only lasts for five years until a competitor knocks it out. No other mRNA LNC product ever makes it to market.

BioNTech comes out $5 billion dollars ahead and almost everything that could go wrong did go wrong.

Now obviously this math is fuzzy and there are ton of variables at play, but I hope you’re starting to get the picture. BioNTech either thinks their entire mRNA pipeline is worthless, or they’re low-balling Matinas into trying to sign a bad deal. From a betting perspective, this is moronic. It’s like arguing with a casino that instead of a slot machine paying out $40/play for a $1 bet, it should pay out $43/play. Just shut up and pull the handle. Make your money and run off into the sunset.

Medium success scenario:

Let’s say 5 drugs succeed and have yearly sales of $1B. Which is a very low estimate given the diseases BioNTech is targeting.

($24B * 20.8%) – ($100M * 79.2%)

($4.99B) – ($79.2M)

Expected value is $4.91B - $500M in milestone payments which brings us to $4.41B.

This is just the first year of expected value with five drugs approved. Now, unless there’s a weird event like a pandemic, most drugs don’t start with peak sales. It takes a few years to build up recognition and getting doctors to prescribe them. Although something like an HIV vaccine would likely trigger global orders somewhat similar to Covid-19.

At five drugs approved with peak sales of only $1B, the expected value is so high it would be like betting a dollar and getting back $44. Per year. For another 10-20 years.

10 years = $441 dollars for your $1 bet

20 years = $882 dollars for your $1 bet

For BioNTech, the expected value looks like this:

Year 1: $4.41B

Year 2: $9.4B

Year 3: $14.39B

Year 4: $19.38B

Year 5: $24.37B

You’re probably starting to understand here why I think that by not pulling the trigger on this the CEO of BioNTech is being objectively stupid.

I am not a scientist. I do not know what the chance of each drug in BioNTech’s pipeline succeeding is. Some of them just say “solid tumors.” But it’s irrelevant. Even other costs like development, salaries, and taxes aren’t worth looking at. It’s pointless to calculate because the expected value is so high.

Remember, these are still ultra conservative scenarios. Just for fun, let’s see what the expected value of a medium to high scenario looks like.

Winner winner chicken dinner scenario

10 drugs are approved and the average yearly sales of each drug in the pipeline is estimated at $3 billion.

($72B * 41.6%) – ($100M * 58.3%)

($29.95B) – ($58.3M)

At this point the upfront payment and the milestone payments are so miniscule compared to the Payout that they’re not even worth thinking about. It brings us to $29.4 billion in expected value, per year, until the drugs fall off patent or are replaced by something better.

This is the realm of deal-making lunacy and we’re only talking 10 drugs with decent sales in an ever-expanding mRNA pipeline.

If BioNTech manages to knock out a drug like Keytruda, which might do $22 billion a year by itself in 2023, then we’re entering all time greatest investment levels.

Which is why it blows my mind that BioNTech is reportedly low-balling Matinas on a licensing agreement. The expected value generated even when most of their mRNA pipeline fails still makes a licensing agreement with Matinas a no-brainer. The upfront payment could be $5 billion, and it still generates positive EV if only a few products in the pipeline are successful.

It’s like arguing over what size of nuke is needed to wipe out a family of 20 gophers.

“20 kilotons will do it.”

“Don’t be ridiculous, all you need is 10 kilotons!”

The answer is, who gives a shit?

An LNC mRNA combination could unlock an aggregate of value that is comparable or exceeding that of Keytruda, the highest selling drug of all time. Which is why it doesn’t matter what you pay for it. Unless Matinas is demanding something like $10 billion then you just pay them what they want and move on with your life.

When making bets, generally speaking, if something has positive EV, you have to take it. Otherwise, what are you doing? Why are you even in the casino?

Did the CEO of BioNTech attend all those meetings in Boston just to waste everyone’s time? If he’s actually trying to low-ball Matinas, as the CEO Jerry Jabour has been hinting at, then I’m convinced that Uğur Şahin does not understand business, economics, or even basic gambling. And make no mistake, investing is a form of gambling.

Maybe I’m off about this. Maybe its BioNTech’s negotiator, or an investment bank that’s leading them astray. But if that’s the case, they should fire everyone involved in this and give Matinas whatever they’re asking for to close the deal quickly. Otherwise BioNTech risks someone else, like Moderna, figuring out how valuable LNC technology is and just outright buying the company.

If I had $2 billion, I’d do it myself. It’s a no-brainer. Given the state of biotech industry Matinas might even sell the entire business for $1 billion. Buying them out is smarter than a licensing deal because then you aren’t paying royalties for the next 20 years.

Right now, it’s like BioNTech is trying to buy a $10M dollar mansion for $100K and then putting a clause into the contract that says they’re going to walk unless the owner repairs a single shingle.

I wrote an article saying that a licensing agreement would be convoluted and near impossible to agree on terms and that BioNTech should just buy out Matinas. Maybe that’s what’s happening here because this is turning into a clusterfuck. It’s been 203 days since the companies signed their research agreement. The licensing deal was supposed to be signed within 90 to 120 days. If BioNTech is going to walk they should just release Matinas from the one-year exclusivity agreement. Let someone with an understanding of expected value and an appetite for risk take a stab at it.

This arena might not be for Uğur Şahin.

Hopefully we get some color on this when Matinas reports earnings on Wednesday. Which for the first time is after hours instead of pre-market. Maybe that means something, or maybe the CEO is just getting old and doesn’t want to be in the office that early.

Thanks for reading and don’t forget to follow us on Twitter.