Hacker News new | past | comments | ask | show | jobs | submit login
Ask HN: How to start a startup that requires a high amount of initial capital?
149 points by break_the_bank 6 months ago | hide | past | web | favorite | 84 comments
I want to start a payments bank in India but it requires a 13M+ USD initial capital just to get the licensing fee.

Wondering if people have any examples of how one might cross this barrier to entry.

There are two ways to answer this question, two new questions basically.

One is "raise over $13m." How? This is not impossible. There is more investor money looking for home today than ever, I reckon. However, this is not a level playing field. Who you are and who you know and how well will make a big difference. This is probably the wrong place for advice about this strategy.

The second possibility is "what is an easier/cheaper idea, that can later become a payment bank?" This is the HN philosophy, from our founder PG.^" Say you had $13,000. What could you build that would get you customers, impress investors or otherwise put you on a less direct path to what you want? Then, you are pitching "turn this thing into a bank" instead of "start a bank" which is a little easier.

This is hacker news so conventional advice here is take on hard engineering problems, and easy business problems. If this was a forum for business people, it'd be the reverse. There, people would rather take the challenge of raising a large sum but avoid hard engineering challenges. You need to figure out what you are strongest at. Are you the kind of person that can cold call your way I to a 7 or 8 figure investment or the kind of person that can build a product in 2 months that will get you investors or customers that will get you to the next step.

Finally, what you are asking is "how do I succeed at this very hard thing." There are no formulas that are easy and will work for most people.

Read this, pay attention to the bridgehead: http://www.paulgraham.com/ambitious.html

very good answer ! (ok_hand emoji, 100 emoji)

I'm 23 and I just started working last year as a software engineer here in London. I don't see myself having a network that allows me to raise $13M in capital.

Taking the longer route to creating a bank while working on something in that direction is really good advice.

I'll read the pg post :)

Yeah, I read all these comments and they miss the obvious answer.

Pitch large vcs and ask for a lot of money.

It goes against the YC grain, but it is the reason there are silicon valley vcs. It used to be that starting a company meant building chips and buildings and assembly lines, and some vcs are structured to write 30m series a, 50m total to build product, and then 50 to scale. A friend of mine is doing this in the database world right now.

If there is a 15m regulatory barrier, then your pitch is why your team is the right one to bet on, and you raise 30m. The money is out there, you can take it

Responding to a few other replies to this thread, you can see funds that give 30mm+ seed/A rounds in Crunchbase or Pitchbook. That’s not to say those rounds are a fund’s specific thesis or the norm, only that it’s possible.

It's worth noting that licenses can represent assets of a company and have value in liquidation scenarios. Someone else looking for a license may just buy your failing company for $11m and get a $13m license.

> some vcs are structured to write 30m series a, 50m total to build product, and then 50 to scale

Could you please give examples of a few such funds?

Are there any concrete examples of these larger scale VC:s and the companies they've funded? The small-fry tech startup is the most familiar story.

Pretty much every large tech company you've ever heard of was VC backed originally. Facebook is a recent example with a couple of the VCs putting in 500 million in a later round (see http://fortune.com/2011/01/11/timeline-where-facebook-got-it...) Also see more recent examples with fairly capital intensive companies like Uber and Lyft. The odds of getting tens of millions of dollars in the very first round is unlikely. But that's also an unreasonable ask by unproven founders: you'd need to show that you have some clue what you're doing with smaller amounts in the earlier rounds... Apple got <$1m from their first investor.

Partner with a bank.

Or get a hedge fund to buy a bank shell, have it as an asset, and use that bank. If your startup fails, they still have the asset.

Or you may just need to have access to 13M, it's probably not the fee. So - you can just borrow 13M and use the 13M as collatoral. If you structure the deal right, it's 100% risk free. Moreover, you may be able to invest the 13M in which case you can still return an interest rate.

So you could find a fund with 13M, draw up the paperwork such that they put that in your company, but it can't be spent, and the 13M is effectively it's own collateral. Then pay the investors interest, plus some small coupon payment, or give them some equity for their risk free investment.

So long as the paid up capital isn't used for anything, it's a 'non transaction', i.e. it has no financial meaning.

That said, the 13M may be designated as collateral or some kind of requirement for the payments system.

> So - you can just borrow 13M and use the 13M as collatoral

wouldn't OP need equivalent collateral to borrow 13M even?

No, because he isn't spending the money. He borrows 13M, and then the 13M just sits in an account, doing nothing, or earning say, the risk-free rate. There's no risk for the lender, because the money isn't going anywhere.

Except if the money gets used for reimbursements/damages that the new startup bank has to pay if it's mismanaged --- that is one of the reasons the money needs to be there, I assume.

No you get an asset worth x and use it as collateral against the loan valued as x. You'd have to pay interest on the loan, presumably from making a basic investment. It depends on what the requirements of that 13M are. But if it's just plain 'paid up capital' ... then the company could take the loan, 'owe' 13M, but then 'invest' the money back in the entity they borrowed it from! The regulators may not like this though. Depends.

So you'd still need an asset worth 13M. What sort of assets are acceptable these days? I ask purely out of curiosity.

Wouldn't "the bank itself" be the asset worth $13m? In the same buy-out as happens to a lot of sports teams, as long as you have friendly lenders, you can structure the deal to buy the asset, using the asset as the collateral (like a huuuuge car loan).

If it's always going to cost $13m to get a bank shell, then (although there might not be many buyers), the bank now has a $13m "thing" which they can take ownership of if it goes down the pan.

True, the 13M isn't the fee but its money that you need to have to start a bank. You don't need to spend it but you need it in case you have some damages/reimbursements.

I'm closely looking at partnering with a bank. I could possibly look at banks that traditionally don't have a lot of debit card sales and see if they are interested in what I have to offer.

I have thought about this myself.

I am from Africa and needless to say, we don't have an abundance of VC funds for startups.

I have used products live Revolut and Monzo before and loved them, there is an opportunity being missed in places like Africa and other developing economies that have a huge growing youthful middle class that hates the way the established banks operate and is comfortable with technology.

I spent most of my free time last year building a backend that can handle some core banking functions, I already did research on the licensing and knew I would never be able to raise that money just with an idea and business plan so to keep myself going and also have something to show potential VC's I decided to start small, things like airtime top-up, bill payments etc

Break it down , look at the bits that you can build out and get some initial customers while you look to raise money, you don't have to go all out and build a whole bank, start small, licensing for things like e-wallets and money transfer is not unreasonably expensive, you can then go on to build on top of that

Good luck.

We are in the same boat. Are you from Africa but working in London/Europe? I am from India but I am working in London.

I have identified millennials to be my main market as well. In India whenever I visited the bank with my father, the person at the bank would try to sell a new card. They always want to "sell" you something, we just have to suck less as a digital bank and the millenials will happily help us get rid of the traditional banks. I've discussed Monzo with a lot of friends, and they have loved the idea.

Curious, what part of Africa are you from? What is the cost of the banking license like there?

Check out Monzo in the UK. They started as a top up payment card with great technology. This meant they didn't need a banking license. They are now moving into the current/checking account space with a license.

So do you need a license from the start? Impossible to tell from your comment.

N26 in Germany worked around the same problem by putting the customers actual accounts into an existing bank and just acted as the front-end for those accounts. They later (1-2 years after launch) got their own banking license and moved everyone’s accounts over.

The idea is heavily borrowed from Monzo. A Payment Bank in India seemed like the closest thing to what Monzo did initially. A full banking license requires 5 times more capital than a payment bank.

That is a good question. I'll see if I can still issue cards that links to one real bank account and if we can bypass being a payment bank.

You can re-issue international debit cards and sell them. The vendor will give you API and you build the interface. If you can convert local indian currency to USD, then it can work for you. From there, you get traction and then ask investors for cash.

I'll look at this. I didn't think of this, I am a little skeptic about Reserve Bank of India allowing this. Thanks for the lead.

I know a bunch of people here have said "partner with a bank" or get a license. That seems like a good option, but obviously you wonder how can a small startup or even a founder or two do that?

I have learned a couple things about being the little guy trying to partner with bigger folks:

- Josh Elman (from Greylock) suggested that when (I think) when LinkedIn (or Twitter, I forget) was small, they didn't go out to look for partnerships with big companies, because they knew those wouldn't pan out because no one in a big company would care. They mostly sought relationships with smaller companies with some audience, but who were more receptive to talk to team up to crush the big guys and offer better value. Smaller competitors in crowded markets like banking are always trying to find ways to become more dominant, so they are more open than the folks at the top of the market.

- Just make your offering really really compelling - the "offer you cannot refuse" kind. This is not easy. This goes against traditional business sense, because you might have to offer value at less than cost. But if this is what the existence or growth of your business hinges on, you have to bend over backwards. Goes to the good old Dale Carnegie stuff; focus on their value in the conversation over yours. They need to feel like working with you is a steal of a deal. There are plenty of ways and business models to do this, but offering human capital, co-branding, offering equity or some rev share, customer insights, etc. are all fair game.

Josh Elmans advice is really great advice. I'll try to contact the smaller banks that don't have a lot of debit card sales.

Coming up with an "offer you cannot refuse" feels a little tough given its just the too of us wanting to do this after working for one year after University.

Start smaller. Most of the disrupter banks started as something else with the view to eventually obtaining a banking license. That way you can actually prove that you have a product worth investing in. Usually banking startups start with a Stored Value Facility license (or the local equivalent) and work up from there.

Seconded. Start smaller. The simple concept to remember is: if you need $13M to start this business, it's a business you can't start right now; it's instead an opportunity for larger businesses.

We had a similar problem: we started with "let's build the city of the future", which would have probably required a large capital (of course, we wouldn't have done it alone, but still...), and ended up focusing on creating innovation at the financial layer of real estate, which we started with a few 100,000s of $$.

Doing good so far.

"The simple concept to remember is: if you need $13M to start this business, it's a business you can't start right now; it's instead an opportunity for larger businesses"

True. I should look at the smallest thing that I can do that can provide value and that can somehow pivot into a bank.

Good luck with your startup :)

This isn't the sort of thing to do as a cold startup. The way to do this is what is sometimes called "intrapreneur", as a "startup" within an existing large corporation that can bankroll such a large project.

Are you an engineer, or a banker? If you don't have a track record in the banking industry, you're hosed. No one will give you that kind of money. If you've actually been there for the founding of a bank before, then maybe. Moreover, if you're an engineer without banking experience, you probably have no idea what you're getting yourself into.

Can you finance the license fee? Ie borrow to get it, and if all doesn’t work out, the corporation goes bankrupt? Or can you lobby the issuing agency to give you a payment plan?

My only concern with such a large investment is that at the beginning you don’t know what you don’t know. You think $13m is a big problem, wait till you have that license and uncover all the reasons why people with $13m in their pocket have avoided the problem you’re trying to solve (just a maybe, but I’m sure there are other problems you’ll encounter you should look to resolve before you try dump so much money on a business).

The fix has two steps:

1) Bring in co-founders whom raised $100m+ from VCs in companies they were in before. Also, one of those companies needed to have had an exit where it was very profitable for those VCs. Getting these co-founders is hard, but it is your fix. One of them needs to have been the CEO of the former successful exit.

2) Use $500k in capital to build the pre-launch version. Be hyper-capital efficient building the tech, for demos to investors. Bring the product working to investors along with those co-founders as part of the team.

Make sure those co-founders confirm that they think they can raise the capital in this business model.

There is a reason requirements to start a bank are so high. Banking is a really hard business and you margins are not that great. There is so much regulation that your licensing fee will look like small potatoes.

Instead you should ask if there is cheaper way to get to where you want to be. You could start by researching how the PayTM guys did it.

The PayTM guys really took off with demonetization and they really rode the wallet wave.

It is critical to minimize the risk. Remember, you are taking somebody's money. Their main concern will be the payout and this is the likelihood of future income minus risk of losses. You probably spent some time thinking of future income (everybody does), now do your homework and work hard how to mitigate risks. With lower risks you can ask for more money for less stake in your budding enterprise.

With large amounts of money involved there will be more focus on risks involved and so it gets critical.

Lots of crazy answers here.

As an industry insider, what you do is buy a bond from a company that specializes in these sorts of licensing bonds. It should only cost you a small fraction of the capitalization requirement each year to maintain.

Another unique suggestion that I couldn't possibly think of. I'm really overwhelmed by the response I got to this thread. I'll look into this.

The capitalization requirement doesn’t exist to make it an exclusive club. It makes it so the potential regulatory fines have teeth because you can’t just fold on the first one received.

The bank offers to cover your fines (with your entire business as collateral), for a fee. You can think of hat fee as the interest rate on a loan, which it basically is because the money is put into an escrow account with the government.

In my own market it was $10M usd required to become a payment services provider (e.g. Venmo). But you could get someone to offer the bond for a few hundred thousand dollars a year, the price depending on how comfortable they were with you and your business. That’s within the realm of doable for a startup.

Find a smaller problem to solve that would require less capital and prove that you’re worthy of $13M.

You can likely do what you want by partnering.

we are a lending startup in India - also a YC alumni (RedCarpetUp YC S15).

We are (now) a regulated entity with specific licenses that took us a crazy amount of time from the Reserve Bank of India.

Two things - first, the business models of payment banks is kind of unproven. Those of us who are in the fintech space in India kind of agree that the regulators kind of screwed up here. I'm hoping you have answers to that. TL;DR - a payment bank cannot lend.

Second, the regulatory structures in India are setup to look at license grants not as a consequence of capital. Even if you had a 100 million USD (and I know of specific examples here), there is very less likelihood of you getting a payment bank license unless you can prove out a gold standard team and board experience. Just to remind you, bank CEOs in India are appointed by the Reserve Bank. They are NOT appointed by founders or investors. One of the CEOs of a very large bank in India just got fired by the regulators for dancing around stockholding regulations - https://economictimes.indiatimes.com/industry/banking/financ...

I feel there is a very low chance of you getting regulatory clearance for this. Indeed, we give out a t-shirt that says "Because, its regulatory" to fresh employees at RedCarpet ! We are tech founders, but we have spent countless hours sitting outside RBI doors trying to make case for several regulatory changes. We were one of the first to get license grants for non-standard lending risk policies to the unbanked in India (that's a billion people!).

Write to me http://scr.im/redcarpetup if you want to come work with us for a while. Maybe you will stay a while and listen to stories of regulatory pacman in India.

I would do a prepaid visa or something.

prepaid instruments in India are regulated. Beyond a certain limit you need to do a full KYC. Issuing a prepaid/wallet needs a license with fairly high capital and regulatory requirements (in the same order of magnitude as the payments bank).

You can issue a card in partnership with a bank or another license holder, but as i said before - in that case you better have a business model for that, because you aint gonna make money from MDR or anything.

Thanks for the response :)

I have read the news about the YES Bank CEO. I was hoping that they would be more relaxed with payment banks.

I'll drop in a mail over the weekend. I saw you guys when I was in my third-fourth year of University. I think the other startup that was doing something similar was Finomena but I see that your target market is different.

Have a lot of respect for the finomena guys. If they had survived, they would have been our top competitors.

You can raise a lot of money at the beginning if the founders have the right connections and track record. Opendoor was able to raise a $10M Series A without a product. [0]

[0] https://techcrunch.com/2014/07/07/opendoor/

The high-entry barrier business is where VC actually makes sense. Building a webapp and getting $3M VC money is a little bit of a waste. Building a rocket or a jet requires $100M and there's no other way around it. You're maybe somewhere in between. Or start small. Maybe in a different country/province, if you can.

Like someone said, partner with a bank. Maybe check with RBL bank, they even have API for virtual accounts (https://developer.rblbank.com/content/virtual-account-api).

Do something in the banking ecosystem to get your foot in the door.

What I have noticed in Malaysia are startups effectively handling the {credit card, loan, insurance} applications & such other Web forms that banks are far too incompetent to do effectively.

Once you do this well, you will be on the right path.

I have similar question too. The initial Capital investment for Storage and Trucks along with Ordering and Distribution Software for a Food Distribution business requires up to $2M+ USD Capital. ( It is consider a large sum in Hong Kong, which has a slightly smaller TAM than say US or China ) While I can guarantee we reach $1M USD monthly revenue within the first 12 months, I have yet to find anyone interested in funding.

If anyone could give any tips or help send me a email.

my username @live.hk

Have you considered starting a payments bank in another country where you can prove your idea with less capital? Then you can use the capital from that to get started in India or if there's not enough, you can use the experience to inspire investments to get the capital.

Also, I hope you're ready for an immense amount of red tape.

I have had the idea of looking at other countries before but I haven't put in time to figure out what countries I should be looking at. Will look into this.

From what I've seen, your expertise and network in the field are what will allow you to raise funds for a capital intensive business.

If you don't already know people and/or have deep expertise in the field, why would someone give you the money?

My thought here has already been said. Start smaller if you can. I'm starting really small with one specific banking issue and the entry fees are much smaller. I'm using some hedge fund ideas to get the capital.

Start a software startup and sell it and use the funds to start this one. Simple!

Consumer or Enterprise? If it’s the latter, get a list of customers who will put in (can be legally non-binding) writing, “If you build X it is worth Y to me.” Get enough of those and take it to a VC.

A team with multiple big exits could possibly raise the money from VCs or self-fund.

More likely, something like this would start inside another company or it would partner with a license holder at first.

I have looked into this area. The best way to do it is to partner with an existing bank. You wouldn't need payment processor license which would likely be very costly.

Must say its a difficult task to convince banks to partner with you.

Have you considered opening a cooperative Bank? You could out-innovate a lot of slow moving small banks. And I think the initial capital requirements are lower.

well one way to 'hack' your way to bankdom is through founding a company offering sketchy financial advice then expanding to sketchier insurances and payday lending, then to banking. takes about 25 years and you get to be public enemy number one doing this. they 'sadly' went under during the crisis though.

because its so difficult most banks are very old and most new ones are established by existing players.

You work with an existing bank. Many examples of fintech startups that started like only getting a banking license much later.

Hey! I'm interested in this goal too. Fancy a talk? Email/twitter is in bio.

You convince an investor to fund you with research (aka expertise)

Can you buy an existing bank that is already licensed?

Still gonna cost more than that $13MM, since that's the asset people are after.

you can rent a licence, there are many banks doing that in EU. I think you need to do a better research..

Can you expand on this? Never heard of this but I'm not super involved in fintech/banking. Something like Solarisbank?

Not OP, but here's generally how you would do it [0].

However, in the EU there are a few other banks to try [1], [2].

In the US, you could try [3].

There's also companies which partner with multiple banking services to provide a cohesive offering [4], [5].

For anyone wanting to do this, here's two issues that I have found.

1) If you don't have banking licenses, the conversation will stop abruptly. Even if they advertise that they may work with you from their website. It's a big hurdle.

To overcome that, I know of 1 company that may help [6]. If anyone knows of any more, please reply.

2) It's rare to talk to VP/Management level operatives, which is where you want to be. Unless you manage to dig out the CEO/COO email and contact them directly. Going from the bottom (contact us form) upwards just means you'll end up speaking to a Z level employee who doesn't know the banking business at all.

Disclaimer: I've contacted over 100 fintech companies. The whole state of affairs is pretty bad. It's improving, but won't be there until 2020 at least.

[0]: https://blog.faisalkhan.com/money-transmitter-license-applic...

[1]: https://www.fidor.com

[2]: https://www.mambu.com

[3]: https://www.crossriverbank.com

[4]: https://www.railsbank.com

[5]: https://www.unnax.com

[6]: https://www.coinx.com

Wow, you know I am always blown away by how helpful people are and how awesome responses can be. I really appreciate this and all the resources, I'm reaching out to Fidor today to learn more


adplexity.com - 5 million USD revenue.

Better launch a new product first, make revenue like the product i posted and use that money to fund your big business.

I know it's not popular around here, but have you considered getting traditional business loans from a bank?

the barrier is there to stop instances like this.

That's uncharitable.

Here in Australia there is a team building a new bank and having good success doing so. In order to overcome the institutional barriers like the one the OP raised, the team consists both of highly experienced corporate "grey hairs" and talented young hackers/designers.

This way, they have the industry experience and network to ensure the company is run competently and prudently, whilst also being able to leverage the energy and creativity of younger hacker-types who can bring new ideas to the market.

I'd suggest to the OP that this kind of approach may work for them.

You just do it

You could try crowdfunding, but I can't for the life of me remember the name of the website. "Kick"-something.

Crowdfunding might be an option if everything else fails. However, it's highly dependent on your network and social visibility.

Disruption. Begging for forgiveness instead of asking for permission. You can see it in many major players that got their start by going around regulations.

Once you are onto something, the regulations can be satisfied by backers who see a business plan, cashflow from the side of the business, or perhaps new regulations will form around the new company's business model.


There is no way around this kind of regulation, and they definitely are not going to 'change their minds' once they see someone has broken it. They're going to put someone in jail.

This works for stuff like uber and airbnb but will land you in jail if you try to do it in the financial or healthcare sectors

That's too big a risk. Its just begging one phone call from a competitor to the authorities to wreak havoc on the startup. This strategy could work in yet uncharted territories (crypto comes to mind) but not in traditional systems like banks in this case.

I actually was thinking more about going around regulations by building a solution which wouldn't be targeted by those regulations, in one way or another (outside jurisdiction or innovative product that does something in different way).

The financial industry is highly regulated and it's a good old boys club; shenanigans like that don't fly and it would land one in jail faster than one could bat an eyelash.

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact