Banking dissonance - Public vs Private

Summary

Either the VCs or the public markets are wrong. Badly wrong.

The two worlds

I continue to be amazed by how detached the public markets and the Venture Capital-driven world are from each other. The valuations of publicly traded companies on the one hand and those of loss-making VC- and business-angel backed growth companies on the other make no sense to me. There is possibly more interplay and information exchange between the public markets and PE (Private Equity) investors, given the occasional delistings and divestments of public companies involving PE. But in the growth company side the dissonance seems to be permanent although the relative valuations between the public-private divide change along the business cycle.

Currently public markets are cheap. I've had a gut feeling for a while that nowhere is this phenomenon more striking than in banks and fintech. So I decided to take a look but did not find reliable data - hence I write about this myself.

For a full disclosure, note that I am board member and shareholder in a small newly created bank Fellow Pankki Oyj | Fellow Bank Plc, but I write in private capacity and mainly out of interest as a business angel who has made the jump to the public markets and tries to make sense of these two worlds.

What does the data say?

The first observation is that data of unlisted neobanks is difficult to find and even more difficult is to make it relevant for apples-to-apples comparisons. To limit the amount of work, I decided to limit myself to a few key metrics and also to focus on three European digital banks (N26, Revolut and Lunar) and one publicly traded bank (TF Bank AB (publ)) that possibly is a bank these challengers might want to be like in the future.

It looks like TF Bank has enviable metrics that the challengers can only dream of, for now. In the table below I have included a few datapoints of the four.

It should not surprise anyone that a publicly traded bank faces pressure to be profitable. But the metrics of TF are very good across the board when compared with the unlisted peers. Capital use is efficient (RoE close to 20% is very respectable in banking business), the growth is very impressive (revenue growth consistently over 20% over the last years) and deposits and lending are largely in balance. Meanwhile, the challenger banks are all in the red (massively so!) and need constant new capital injections. But for a bank it is also very worrying that the loan book is very small compared to deposit base.

In a rational world and indeed within the public markets, high profitability tends to be rewarded with high multiples. Highly profitable and well managed banks like Santander are given higher P/B (Price to Book) multiples by the market than the European laggards like Deutsche Bank.

Not so across the public-private divide. The company (TF Bank) with best metrics is easily the cheapest on P/B valuation and naturally is the only one with a meaningful P/E multiple.

Now, my data should be treated with caution. Unsatisfactory for digital and supposedly agile companies, N26 and Revolut have not yet published full 2021 numbers, forcing me to use the latest 2020 data. In addition, both companies have their banking operations in a subsidiary and the consolidated figures are of a non-banking entity. They have also raised new cash recently, hence not visible the 2020 financial reports. But none of this would change the conclusion: Loss-making private companies are very expensive. What explains this and what do the VCs see here? Are retail investors better off with public markets? Do IPOs of the neobanks ever make sense?

It's balance sheet, stupid!

The VCs always look for scalability and high growth with improving margins. In banking, things might be different.

How to make the balance sheet work? During the negative central bank rates, a major part of the excess deposits of the challengers had to be parked in central banks, where they earned negative interest rates. This problem is now gradually disappearing, but the most obvious remedy and indeed traditionally the raison d'etre for any bank has been to lend out the deposits.

It is notable how different strategies the challengers have adopted. Some are clearly aiming to increase the balance sheet business and of our examples N26 has belatedly talked most clearly about this while Revolut for a long time did not seem to care about the excess deposits. Rather, they chose the super-app strategy, with some success - the fee income boosted revenue to a respectable percentage of the balance sheet total. Lunar has not really monetised their balance sheet or client base this far.

But for the balance sheet business to work, additional capital is constantly needed. N26 recently said that they wish to reach profitability without extra capital. For me, this would suggest quite limited growth, if the profitability is to be reached with lending. It is likely that they will try to imitate Revolut with more fee-based income, e.g. from stock trading and cryptos. Revolut is more advanced in fee-generating business but they have an even bigger problem in making their deposit base work.

But even in the best possible world, where the two would consistently exceed the 20% RoE level TF Bank exhibits, P/B levels of them can hardly be justified as banks. A better strategy probably is to find new ways to generate business that is not dependent on the balance sheet and capital requirements. This could get the exit valuations to the level of public peer company Adyen that has a scalable platform economics and no banking licence. Despite the recent market turbulence, Adyen has P/B multiple closer to 100 than 1.But then again, Adyen is profitable.

Conclusions

Based on this little exercise, I remain very happy with my choice to focus on public companies instead of unlisted equities. When the VC backers of the neo-banks sink billions into their ventures, there is little hope for a business angel to get a stake in these companies anyway. But for what I can see, investing via the stock market may be a better choice anyway.

Comments welcome!