Hacker News Reader

Can you go into a little more detail about who owns the bank account that users are depositing into when they use sparkswap?

AnchorUSD is our payment partner who operates the API we interact with (it's a custom API that they built for us).

The escrow account is held with a trust company in AnchorUSD's name, but is designed so that only the users they onboard can access the funds. Funds cannot be legally used for any other purpose.

> The escrow account is held with a trust company in AnchorUSD's name,

That means the total FDIC insurance on all the money in all accounts is capped at $250K between all the users of your platform. One Bitcoin quoting at $10,288, means that at most money needed to buy 24.3 bitcoins are insured across the entire platform.

Personally, I don't expect FDIC insurance on swap space used for trading, or generally anything except my bank account.

It is not about trading - customer's funds are deposited into a bank account of a US financial institution partner under under a name of a trust created by their service provider. Until those funds are "spent" they sit in that account. That account is FDIC ensured to $250k.

Compare this to you depositing money with your brokerage to trade stocks. The account is opened at clearing broker in your name, which means that the SPIC coverage limit is just yours.

Our partner has clarified that FDIC insurance is per user (per Tax ID) rather than for the whole platform

I'm sorry, but your response is a strong indicator that anyone and everyone needs to stay as far away as humanly possible from your company in its current form because:

1) company's leadership does not understand how FDIC insurance works -- trusts have their own tax ID which are not customer's tax id. Trusts cannot currently be created in real time.

2) company's leadership does not know what they are getting from a "partner" -- some random person who ran a couple of companies and consulted for some financial institutions should not know more about what is and what is not possible than a leadership of a company in a space

3) there's a lack of understanding among the leadership what are the implications of customer account being insured by FDIC.

4) leadership does not understand that ACH credit has a liability attached to it which needs to be managed very carefully ( especially with points 1-3 )

5) In no even should the points 1-4 stand should a company ever trade from its own inventory to satisfy customer's order.

Drop whatever you are currently doing and immediately hire a compliance person from a large commercial bank that does not have any reason to leave his or her position i.e. overpay for one of 100-200 experts that exist in this field so you can at least ensure that you don't blow up in the most spectacular fashion with a potential criminal prosecution.

This is not an Uber or AirBnB. If you mess up you will create potentially criminal problems not just for yourselves but for your customers.

thank you for making this public, even though it may not be perfect.

So I need to trust

1.AnchorUSD

2.Their trust company where escrow is held

3.Zap for lightning network wallet

4.Abacus for identity verification

5.and you (sparkswap)

I'd say this is neither direct nor instant (just abacus verification adds crazy friction). And the trust issue is real, I'd personally feel better with one large custodial institution than a number of startups in this chain that I know nothing about (or their security practices).

If you can solve that there is market for sure.

Initial setup is definitely not as fast as we'd like it to be given that we have to do identity verification, both for fraud prevention and AML compliance. Once you've gone through that setup, the interaction is near-instant, and the Bitcoins go directly into your wallet.

Both Zap and the Sparkswap app are open source, which should reduce some of the trust you have to place in them.

The trust issue with fiat is difficult to get around, with both Sparkswap and any other service that interacts with the traditional financial system. By removing custody of at least the Bitcoin side of the transaction, we think we're taking a step in the right direction.

Overall your feedback is well-taken - there is certainly more we can do to reduce the number of steps and different pieces of software / service providers that users need to interact with to use Sparkswap.

> I'd say this is neither direct nor instant

You're correct in other ways as well:

The "lightning" network (heh) still requires slow Bitcoin transactions to initiate and close. And it's between two trusted parties anyway, so this negates the need for a trustless network like Bitcoin.

A far better "lightning network" has existed for many years. It's called a shared SQL database.

No it's not between two trusted parties. If those parties disagree, they have recourse to the blockchain, and so don't have to fully trust each other.

The lightning network couldn't be built without having a lower level trustless party. Bitcoin actually lets us bootstrap trust.

More thoughts on bootstrapping trust:

https://www.nemil.com/crypto/why-cryptocontracts-matter.html

Or centralized operation with decentralized checking that can be implemented with SQL databases and/or logs. So much more efficient. One can use non-profits/foundations with the protections built into their charters, policies, and contractual agreements with steep penalties for violations. As in, use existing tools with tons of operational experience and case law supporting that they work. Also, these have basically none of the drawbacks of the fully-decentralized solutions.

How do you handle ACH clawbacks? What your business models needs to protect against is... hackers steal a victim's private information enough to be able to submit the ACH to you. The victim reports it to their bank several days after the theft. The bank does an ACH clawback from your banking partner.

Where is your time gap to get past ACH clawbacks?

We have guards in place to protect against fraud by verifying your identity in a few different ways, but this is something we'll monitor and make improvements to over time. It's critical to the survival of the business that we have strong fraud detection methods.

However, we also think that the experience of instant transfers is valuable, so on a risk-based approach we allow instant transfers using ACH for certain customers and certain transaction sizes.

> However, we also think that the experience of instant transfers is valuable, so on a risk-based approach we allow instant transfers using ACH for certain customers and certain transaction sizes.

You have a fundamental misunderstanding how ACHs work. There's no such thing as an ACH that is "completed". At +180 days the state is "It is probably going to be fine". At +365 days the state is "It is probably most likely going to be fine" etc.

Source: Done consulting for execs at medium sized banks.

You're right that ACH reversals can come far after the payment appears to be settled. Based on the customer and transaction size, we treat the ACH as "completed" (i.e. available to buy Bitcoin with) at different points in its lifecycle.

You might have a fundamental misunderstanding of their business model.

They don't need to classify an ACH as "completed" or "not completed", they only need to place a confidence interval on its completion state, which becomes easier as they process more transfers.

> They don't need to classify an ACH as "completed" or "not completed", they only need to place a confidence interval on its completion state, which becomes easier as they process more transfers

They ignore financial regulations to the level of Uber ignoring taxi regulations. The difference between the taxi commission or whatever its name is in San Fran and US Department of Treasury is that the later relies on people with guns and can block customer's ability to do any financial transactions.

The reason why in the standard custodial scheme ACH transactions are fine ( but still are frowned upon ) is because the appreciating asset can be reclaimed from a custodian via courts.

All legit exchanges (e.g. Coinbase and Gemini) have this problem but there's really no better way for consumers to transfer USD.

Non-transactional nature of ACH are a feature, not a bug. The legit exchanges account for it using in house compliance departments, risk management groups and people intimately familiar with the cash management side of the finance industry.

Any plans to make this less technical and more user friendly (like users not knowing even about lightning network). I think this has a lot of potential if you make it more mainstream by abstracting away the tech stuff.

Yes, that's definitely a goal of ours over time. We're starting with the existing Lightning community, but we plan to continue to reduce the technical requirements so that users don't need to know anything about Lightning to use Sparkswap.

Great to see this, easy onboarding to Lightning Network is needed and this solves a real pain point with added benefits.

what would you say are the real pain points that are being solved here?

Namely that Im not using coinbase, who are essentially the facebook of the crypto world.

How do you know AnchorUSD, Zap, Abacus, or Sparkswap are not doing the same as Coinbase either?

............

how do you make that leap exactly? and how do you think Sparkswap is going to be any different than coinbase?

aka, buy bitcoin instantly once you've deposited your USD into some partner bank account which you don't have access to but Sparkswap absolutely does.

This isn't really new or interesting and has been shut down repeatedly before because most crypto companies can't handle the fraud mitgation, here's a clue, it's not cheap or easy and can't be automated at this small scale. Instant access to the coins has never really been a boon to anyone other than scammers. The loss of private keys is a huge setback to the consumer and to adoption, which is why Coinbase has been one of the most successful applications within the crypto space.

I don't mean to say that this isn't something idea and shouldn't be done, its just not really interesting or useful.

buuuuuuuuttttttt, then you had to go and say this: > a real product that solves a real problem and enhances the biggest proven use case in the cryptocurrency industry: buying Bitcoin.

the biggest proven use case in cryptocurrency is... buying crypto...?

I've been in crypto full time for the last 6.5 years and that statement actually makes me want to quit.

> I've been in crypto full time for the last 6.5 years

Do you really believe the "coins of today" are going to stand the test of time as hardware gets more powerful in the next decade?

Where do you personally see crypto in 10 years in terms of real world usecases that aren't fraud / money laundering?

> Do you really believe the "coins of today" are going to stand the test of time as hardware gets more powerful in the next decade?

Bitcoin sure, because it actively gets less efficient as computers get more efficient. The only thing that'll stop it getting less efficient is a grey-goo like scenario where it consumes all the world's power. Or, I guess, if we run out of leading zeroes.

I'm surprised to see this misunderstanding persist today, though it was one I frequently saw in 2010/2011.

Bitcoin's mining difficulty automatically adapts both up and down to keep blocks at roughly 10 minutes, whatever the offered resources are. It doesn't monotonically become more difficult.

Of course, I never meant that it couldn't get easier. It just won't for as long as there's demand for it. As long as there's money to be made in mining, resources will be thrown at it. As more resources are thrown at it, it gets less efficient. As resources improve, it becomes less efficient. The only thing that would allow it to get more efficient is if nobody cares anymore, as happened across the lions share of 2018.

Nothing about this graph suggests we should expect anything else [1]. The perceived misconception seems more like a technicality :)

[1] https://www.blockchain.com/en/charts/difficulty?timespan=all

I want to say it means more like the on ramps to crypto are kind of fucked. Coinbase and other on ramp providers have terrible fees, its almost exclusively the primary reason I don't swap fiat into crypto currency.

Those fees exist to cover risk mitigation - i.e. a certain percentage of on-ramps are fraudulent and the fee disperses the loss spread to everyone who is on-ramping.

Unless this service has invented some new technology in fraud prevention, it's going to eat that risk at a loss to themselves.

Poloniex has zero fee wire transfers (other than the wire fee you pay) into USDC which you can then trade on Polo with (paying fees) or withdraw the USDC anywhere you like that has better fees.

IIRC fees are around 0.25%; what do you want to do with crypto that you see such fees as a problem?

I definitely agree that fraud mitigation is the key factor in a business like this succeeding or failing. Instant access to coins is certainly something that can attract the wrong type.

There are certainly pros and cons to self-custody - it's not for everybody. Given the value proposition of Bitcoin and the huge amounts of cryptocurrency stolen from exchanges, it's important for many people to hold their own keys. We're like that ourselves: we didn't want to choose between holding our keys and still being able to trade, which is part of why we built Sparkswap.

why don’t you just one of the many other providers?

> the huge amounts of cryptocurrency stolen from exchanges, it's important for many people to hold their own keys

Can you cite the data for your claim that there is less loss of Bitcoin from holding keys vs an exchange? I'd image the number of Bitcoins lost people people don't have their keys is large.

> This isn't really new or interesting [...]

The list of HN posts that this comment has been applied to is almost as long as the list of successful YC startups.

Wow, this was seamless. It auto-detected my lightning wallet, and the purchase was instant.

As far as I know, this is literally the only way to get Bitcoin on lightning directly with USD.

I have had good experience buying bitcoin through Square Cash app. It can then immediately be sent to the walet of my choosing.

Congrats on the launch Trey! I'm very excited for the future of Sparkswap and the Lightning Network ecosystem.

Shameless plug: I've built a platform that showcases apps that are powered by the Lightning Network.

Most of them are experiments with instant settlements and micropayments, but it should give you a good idea why you should care about Lightning: https://www.lapps.co

Any plans to support the opposite feature: selling Bitcoin over the Lightning Network?

Yes absolutely! Enabling selling Bitcoin is near the top of our list of priorities.

First of all, this is very neat. I just tried this with $15, but I'm receiving the error message "your account must be approved prior to trading". At the same time I received an email from AnchorUSD stating "We expect your USDx to post to your account by September 25, 2019". Does it really take 6 days for the funds to be available in my account, or am I misunderstanding?

I think we're working together in our support channel, but the email you received was in error - we're working to resolve.

Yup, I'm all setup now. Thanks for the help.

Very interesting, i cant find neither in your post nor your website, how you are winning money. (and how much are the fees)

Good question - we make money by selling Bitcoin for slightly higher than we buy it from a few sources. We aim to have very competitive prices (lower than many places you can buy Bitcoin today), but we don't have a set fee rate that we publish.

So if I understand correctly, the business model is to be the only seller in this exchange, with a profit margin justified by the convenience of instantly available Bitcoin funds.

Would it not be possible for a competitor to undermine this business model by running an exchange (based on the same technology, perhaps the same implementation) where selling is more permissive?

When we add Bitcoin sales, it will look more like what you're describing. We've experimented with an open orderbook model in the past, and we're not opposed to it, but we've found that the UX of a listed price is better for most users.

Why are Coinbase fees 1% or more while Coinbase Pro fees are 0.25%? Because noobs don't want to see an order book and hiding the order book from them is worth real money.

This looks promising. Looking forward to trying it out! I do wish there was more information on the website about how it works, fees etc.

I am admittedly a crypto currency novice, so I'm not real familiar with the lightning network and how it works. I am curious how your product differs from the Bitcoin feature on Square Cash app?

I have had a good experience buying Bitcoin with USD from Square. It's fast, there is no fee (although they pad the price a little bit), and I can then immediately send the coins to the wallet of my choosing.

Seems like your product is very similar. Are there any major differences? What are the reasons to use your app and not keep buying from Square?

The product is very similar to the Square Cash app. The main difference is that when you buy Bitcoin with Sparkswap, the Bitcoin goes instantly (and without paying on-chain fees) into a wallet for which you hold the private keys (rather than a company holding the keys on your behalf). And that Bitcoin is available for spending on the Lightning Network.

The downside is that you have to be comfortable being responsible for the keys, which not everyone is.

Great, thanks for the information! Depositing straight into my own wallet is very appealing.

I'm just starting to get into buying crypto, but for someone like me, I'm ONLY interested in managing my own private keys.

I don't typically buy large amounts of bitcoin at a time, so when I do it is advantageous to reduce the fees as much as possoble.

Excgange fees and/or network transaction fees can take a painful percentage from small transactions, so I will certainly be happy to give Sparkswap a try!

Is there any cost reduction of using this method, compared to buying bitcoin from a regular exchange? As I recall taking money in and out of the LN channels still costs the blockchain transaction fee.

If you close a Lightning channel (or use a service provider to translate a Lightning payment to an on-chain payment, a common way to "withdraw" from Lightning channels), you will incur a blockchain transaction fee, similar to using regular exchange. You'll have more control over that fee based on your time preference, and importantly, you won't need to incur that fee to make sure that the Bitcoin is actually in your possession. So it will likely end up about neutral cost-wise (maybe slightly cheaper?) but you'll have more control over the whole lifecycle.

You do need to pay to both open and close a channel, so you'll incur two transaction fees in contrast to one with a regular exchange.

Sparkswap opens the channel to the user, so Sparkswap will incur the channel opening and closing fee.

That's nice. So similar to how a regular exchange pays for your withdrawal transaction. (And of course if it's better or worse depends on how much they'll charge).

One advantage is that your funds are already in a lightning channel, so you can immediately use it for payments within the lightning network with very low fees (vs having to pay onchain fees).

Given that the average computer user is not very tech savvy, and is likely to succumb to malware and drive-by hacking attempts, what makes you think it's safer to encourage people to be their own custodians?

Or doesn't it matter at that point, because you're off the hook?

It's true that this isn't ready for the average computer user. But there are projects that I do think make it safer, for example dedicated hardware devices like the Casa node.

I think that as this technology (Bitcoin and Lightning generally) matures, we'll get better tools to make it safer for average users (just as we've done on the web).

Congratulations on launching!

How are you handling AML/KYC? Is this something that AnchorUSD will front first, and is there a fee that they take for the deposit? Also -- do you have to restrict usage for certain jurisdictions (e.g. NY state residents)?

We have an AML program in place and do identity verification on users that want to buy Bitcoin. We are restricted to 22 states in the US right now (which excludes NY state) but we're expanding that number rapidly, and hoping to expand internationally in the coming months.

There are no fees for deposits.

Do you have a list of states you operate in?

This company will mess this up at some point (user data breach, or lost funds). Just wait until Bisq integrates with lightning at some point. Be patient, wait for, and if you can, contribute to, the open source version.

What is an ACH?

And when you say "Connect your bank account..." does that mean this only works in the USA? Would be nice if this was mentioned somewhere.

ACH is a method for sending money between banks primarily in the US. Yes, unfortunately we're limited to 22 states in the US right now, but we're working to expand to more states and internationally in the coming months. We're working on publishing that information more broadly, thanks for the feedback.

A good cryptocurrency exchange-https://365cash.co

This is awesome! Congrats! Self custody + lightning out of box is a winning combo.

will you also make an android application available?

Sad to see so much talent continue to go down the Bitcoin toilet.

> for the first time, you can have fast, convenient trading without custodial trust.

Really? From your description it sounds like I would have to trust Sparkswap and some other unknown bank with my fiat.

> In just the first half of 2019, almost $500M^3 was stolen from custodial exchanges.

And why would you be immune to this?

> product that solves a real problem

What problem? To buy LN coins instead of Bitcoins?

These are valid questions, not sure why you are being downvoted.

[replying because I previously downvoted and another respondent didn't know why people were downvoting.]

> Really? From your description it sounds like I would have to trust Sparkswap and some other unknown bank with my fiat.

Fiat transactions have extremely high reversibility often for months after the fact, and law enforcement is actually responsive around fiat theft in a way they haven't been around cryptocurrency theft. As a result, at least so-far, exchange thefts have not centered around the fiat side of the operations, and in the rare cases fiat has been lost it seems to have always, AFAIK, been cases where the exchange self-traded in order to cover up a cryptocurrency loss. But perhaps as the cryptocurrency side gets derisked we'll see more fiat shenanigans.

>> In just the first half of 2019, almost $500M^3 was stolen from custodial exchanges.

> And why would you be immune to this?

They'd be immune because those thefts targeted huge standing balances of highly liquid, irreversible cryptocurrencies, that users left on exchanges for convenience in trade. It sounds like their intent is to eliminate all large standing cryptocurrency balances. I have no idea if they will be successful, but if they are it should be a substantial derisking.

> What problem? To buy LN coins instead of Bitcoins?

There is no such thing as "LN coins", lightning is a bitcoin smart contract that allows effectively instantaneous transfers of Bitcoin between users. It works by telling the bitcoin network that some collection of your coins will be governed by a programmatic contract between you and another user, then you can trade your coins back and forth as fast as you want at no cost, and you only need to go back to the network if there is a dispute between parties (where the network then forces correct handling) or you want to remove the coins from the agreement.

This is conceptually similar to how you enjoy the protection of the law and the courts while virtually never going to court yourself-- except more powerful because taking your dispute to the network is completely automated, radically cheaper than a court, and not so subject to making a wrong decision. :)

(The network part of the name comes in because the lighting contracts allow transitive trade, e.g. if Alice and Mallory have funds subject to lightning contract, and Mallory and Bob have funds subject to lighting contract-- then Alice can pay Bob, transitively via Mallory without posting transactions in public. This is possible because the protocol is constructed so that multiple trades can be made cryptographically atomic-- where either all succeed or all fail.)

Part of the reason that you're being downvoted is because misinformation claiming that lightning issues some kind of additional or alternative cryptocurrency is a dog whistle of some well known cryptocurrency scammers, such as Craig Wright and his staff.

Downvoters may have been assuming that you were employed by one of these scammers, but it's just as possible that you simply picked up some of their terminology in ignorance.

Cheers.

Readers should now this is one of the main Bitcoin developers behind the "fee market" and in 2017 celebrated $50 fees, caused by his ideas.[0] While being the CTO of Blockstream, a company whose business plan is to create side-chains like LN. By keeping on-chain fees high they managed to artificially create a need for side-chains, which they could fill with LN and recently Liquid.

Anyway that's a lot of words to not answer my questions other than guessing their intent to eliminate large cryptocurrency balances (using LN means they do have to hold Bitcoins).

I used the term "LN coins" to signify that LN have fundamentally different properties than Bitcoin, whereas your wall of newspeak suggest LN and Bitcoins are the same. They are not.

LN operates by trading unconfirmed Bitcoin transactions, and holders are expected to challenge fraud not to loose their coins. Yes this mean you have to be always online (or you need someone else to be for you). Also you might not be able to pay someone, because there is no route with sufficient funds. There is something called the "inbound capacity problem"[1] which limits how much you can receive.

In fact LN developers are saying that LN is not ready and should not be used in production, simply because it's not ready. See for example a recent CVE which has been exploited in the wild.[2]

Implying, as you do, that using LN is as good as using Bitcoin is far worse misinformation than you misinterpreting my choice of words as if I said LN is another cryptocurrency. (Also funny how you react to that, as if all cryptocurrencies other than Bitcoin are scams, which is of course not the case).

[0]: https://lists.linuxfoundation.org/pipermail/bitcoin-dev/2017...

[1]: https://blockonomi.com/lightning-network-inbound-capacity/

[2]: https://lists.linuxfoundation.org/pipermail/lightning-dev/20...

Transitioning from subsidy to transaction fees to reward miners existed as an idea before nullc. It's in the whitepaper for crying out loud.

>Once a predetermined number of coins have entered circulation, the incentive can transition entirely to transaction fees and be completely inflation free.

At any rate, the rest of your commentary is inaccurate. Lightning channels can go offline and be completely fine. I've had one offline for months before I brought my node back online and everything was fine.

The inbound liquidity issues are being resolved through dual-funded channels, which means both parties of a channel coordinate to start a channel off with the liquidity balanced between them instead of completely on one side.

LN is in fact still in beta, and you should expect bugs like any other beta software. That should come as no surprise.

Lightning is however in a state where it is more or less safe to use. The twitter bot I run that relays Blockstream Satellite transmissions has on order of 3,000 tweets -- which means there have been 3,000 payments with Lightning for just this one service.

Keep in mind Lightning in general is more suited to small amounts given the very cheap fees and divisibility down to 0.00000000001 of a bitcoin.

So to sum it all up, your post is pretty inaccurate.

> Transitioning from subsidy to transaction fees to reward miners existed as an idea before nullc.

That's a completely different idea than having full blocks where users compete for fees. You can also fund miners via lot's of transactions instead of few transactions with large fees.

> Lightning channels can go offline and be completely fine.

That's only because the other party didn't do anything. It's like saying leaving your door unlocked and open is safe because you did it once and nobody walked in to steal your stuff. You were lucky.

> Lightning is however in a state where it is more or less safe to use.

I did post link to a recent bug where LN is exploited to steal users funds didn't I? This kind of misinformation that "LN is more or less safe" really downplays the existing issues. It is beta software and should be treated as such.

> So to sum it all up, your post is pretty inaccurate.

To sum it all up, I was saying transactions in LN are different than Bitcoin transactions and that LN is only in beta, which you also confirm.

Launch HN: Sparkswap (YC S18) – Buy Bitcoin Instantly over the Lightning Network