Possible solution to custodian loss of value due to BTC exposure

It shouldn’t be news to anyone here that our custodians are taking on large amounts of risk by being exposed to NBT/crypto trading pairs, so I won’t go into the details of that here. What I would like to talk about is a possible solution that I have been discussing with Ben.

By using BTC leveraged futures contracts, like found at Okcoin.com and 796.com, custodians could effectively hedge their entire BTC holdings with an account made up of only a small fraction of their BTC reserves. This could be managed in such a way that no risk is taken on for any purpose other than breaking even. Here are some stats and a couple of examples …

1 Okcoin BTC contract = $100 worth of BTC
1 Okcoin BTC contract, at 10x leverage, costs $10 plus a small fee.

Example 1:
Custodian holds 100 BTC
BTC market price: $200
Value of BTC held by custodian: $20000
The custodian takes out 200 short contracts to control $20000 worth of BTC. Over the next few hours, the value of BTC rallies to $210. The value of the custodian’s 100 BTC is now $21000 and the value lost from the 200 short contracts is approximately -$1000, causing the custodian to effectively break even.

Example 2:
The next day, our custodian sees that 20 BTC have been sold for NSR, so she closes 32 of the short contracts.
Custodian holds 80 BTC
BTC market price: $210
Value of BTC held by custodian: $16800
Short contracts held: 168
The value of BTC then takes a dive to $190. The 80 BTC are now worth $15200. The short contracts have gained approximately +$1600 in value, effectively breaking even.

These are rough examples and I don’t think my math is perfect but they should get the point across. If this could be scripted or incorporated into NuBot so that contracts could be added and subtracted (through Okcoin api) as soon as custodian BTC holdings changed, we could effectively stop or drastically reduce the losses taken by exposure to NBT/BTC pairs. This would not require any special trading knowledge or market prediction. It would simply require that the bot know how much BTC is held and how much risk the leveraged account could take to avoid margin calls.

8 Likes

This helps but the exposure to exchange default risks is still significant.

Can someone point to the post several days ago where someone proposed a method to trade between NBT and cryptos with no exposure to anyone? It needs the exchange to execute extra steps. I want to re-read it but can’t find it now.

This is a great idea that should be evaluated more closely. The most important question is what is the cost of this insurance? Perhaps historical charts of BTC and contract pricing can give us a good idea of that. Once that is understood, we would be in a position to discuss doing this on a pilot basis, which if successful, could expand to being used for all BTC held by Jamie And KTm. Whether LPCs providing their own funds use this type of insurance is really up to them individually. They just need to decide if they wish to speculate on the price of BTC. We should avoid speculation with shareholder funds where practical, however.

Whether such an action requires shareholder approval as an amendment to existing contracts is a grey area. I would say if it doesn’t seem to create controversy here in the forum, formal shareholder approval is not needed. Whether it creates controversy will largely depend on the cost of this insurance.

@pennybreaker, if you are interested in taking on the task of assessing the cost of this insurance, I would encourage you to do so and bill for it.

Edit: It would also make sense to consider hedging the BTC that comes from NuShare sales, which currently is about 350 BTC. We need to keep some of the NuShare sale proceeds as BTC so that they can function as tier 4 buy side liquidity, which has been used on several occasions to make raising interest rates unnecessary.

2 Likes

smart.

Can we make some offline-tests to prove the concept, before deciding whether including it into NuBot?

I can work with @woolly_sammoth to write some wrappers around OKcoin futures API, and write some example programs. Would you be able then to make some tests with them?

1 Like

I like your idea and it has inspired me to think about that as well.
A possible loss is not only existing in BTC trading pairs, but other trading pairs as well (at the moment PPC).
The idea that crossed my mind is not only useful to help prevent losses by price decline of funds held in crypto currencies, it also helps to mitigate other issues.
Basically it is a derived version of the several tier system that was proposed here:

The idea is further based on my perception that the sold NBT are ultimately backed by NSR.

What if the BTC, PPC or whatever is bought for NBT would be traded in parts for NSR by the NuBots (and back if need be)?
The advantage is mitigating the risk of declining BTC and PPC price by shifting it to NSR.

Tier 1

A part of the NSR is kept at exchanges and gets traded for BTC or PPC in case the sell walls get thin.

  • the risk of the exchange being attacked remains
  • the risk of the custodian being attacked remains
  • the risk of an exchange minting with the NSR remains

Tier 2

A part of the NSR is moved from the exchange to custodian owned wallet. In case the BTC or PPC wall gets thin, they are moved to the exchange and traded to support the sell wall.

  • the risk of the exchange being attacked is mitigated
  • the risk of an exchange minting with the NSR is mitigated
  • the risk of the custodian being attacked remains
  • the risk of custodian minting with the NSR is created

Tier 3

A part of the NSR is moved from the exchange by a multisignature transaction.

  • the risk of the custodian being attacked is mitigated
  • the risk of custodian minting with the NSR is mitigated

Identified risks:

  • depending on the trading volume the walls might not only grow thin, but may be empty as well. This is no completely new risk. It already exists, because the BTC and PPC in the control of custodians could at the moment be sold for way less NBT than at the point of time when they have been bought.
  • this is no new risk, although it has a new face.
  • as long as the NSR volume at exchanges is low, the NSR buy/sell activities move the market

The financial risk is (depending on the part that is traded to NSR) shifted to the risk of a fluctuating NSR value - a shift that might be considered.

Remark about the NSR in custodian wallets:
NSR in cistodian wallets need to be prohibited from being used for minting.
They don’t represent an active part of Nu in this special case.
So they mustn’t be allowed to be used for casting votes.
Security wise this is not bad for the Nu network at all (as long as they are kept safe - that might be one of the issues), because this effectively reduces the amount of NSR that are available for minting and hence the security!
As long as all other minting goes on, it effectively increases the mint rate!

1 Like

Is there a futures market for PPC? I recalling hearing about one at some point in the past but I’m unclear if that was just talking about a hypothetical market or if one really existed.

That is an interesting idea. I have been buying some NSR with PPC when PPC would start to decline.

1 Like

It might be interesting, but it for sure has flaws and challenges that need to be faced.
I wanted to post the idea anyway.
I have no solution for tier 3 multisig, because I don’t know how to define or elect people that have the keys.
One idea that comes to my mind would be “Tier 3 custodians”, but I’m aware that this introduces a completely new risk.
Another idea is being able to burn those NSR and create them if need be, but a motion to create those NSR might take too long and such a kind of mechanism is not implemented as far as I understand Nu.
The beauty of the latter idea is that this way the value of the traded BTC and PPC is redirected to the Nu network by burning the NSR and decreasing the total amount of NSR.

So if there is a loss in value in PPC or BTC, ostensibly Nu has become fractional.

Isn’t the simplest solution to burn an equivalent value of NuBits to get the Assets-to-liabilities equation back in line?

And in regards to burning — instead of the NSR/NBT exchanging concept, why can’t we take the simplest approach of having a custodian burn NuBits by sending them to a dead-end NuBits address, or making a NBT-to-NBT transaction with a whoppingly excessive transaction fee.

Am I missing something beyond this very obvious approach to burning?

Perhaps part of the role of a custodian will be to burn NBT as directed thru votes / motions. (Can a custodian vote have a negative value for NSR qty?)

You have to burn NBTs that are in circulation. To get NBTs in circulation you gave to buy them, Buying with NSR is the obvious way.

No, because what’s being described is a different problem. Here’s a simple scenario:

1000 NBT are in the sell-wall on the NBT/SomeCoin market, priced at 1 NBT/SMC

500 NBT are purchased for 500 SMC

Tomorrow, the price of SMC drops in half, to 0.5 NBT/SMC. The previous buyer wants to sell their NBT back for SMC. Unfortunately, due to the price drop the buy-wall only contains 250 SMC.

Burning NBT at that point isn’t useful. The NBT holder wants SMC, not less NBT in the market. That’s why the OT is describing the use of a futures market to hedge the custodian’s liquidity.

@JordanLee
The fee on 796.com is 0.03%. The fee on Okcoin.com is 0.03% for opening contracts and 0% for closing contracts. Spreads on both sites tend to vary between $0.10 and $1.00, being on the lessor side when trading activity is quiet. Liquidity is good on both sites. Placing limit orders at the last bid or ask price tend to be picked up fairly reliably. I’ve experienced outages and 2fa reliability issues on 796.com. Okcoin seems to be the better site.

It is also worth noting that both sites require the user to ID themselves by providing phone number and passport or other government issued ID number, so it will be necessary for each custodian who is interested in taking this route to setup their own private account. The phone number does get verified. I’m not sure about ID number verification, but I have a feeling that this is only verified when fiat is deposited or withdrawn.

@desrever
By offline-tests, do you mean something like paper trading?

Yes, I would be able to help test the API and example programs and consult with you on use of the futures exchanges.

@masterOfDisaster
I really like your idea. In theory, it would be a solid option. In current reality, NSR price spreads and low volume would make for incredibly unreliable and costly order fills. What would be quite useful but kind of a double-edged sword, is a strong competitor with a similar product who’s peg we could hedge against. :wink:

@Ben
I don’t know of any PPC futures exchanges, but if they exist, I doubt they would have the volume necessary to cost effectively use as a hedge.

@Excoin
Did I hear something about NSR futures exchange? Because that would be sweet! :smiley: Let me know if there’s anything I can do to help you guys with concept or testing. I’m VERY interested.

Yes, I’ll be able to spend time testing different scenarios for/with you.

Great idea. However, I recommend considering BitMEX over okcoin or 796.com (partially due to trustworthiness and partially due to their available features). Clawback can be painful!

I see bid/ask spreads of +$15 on that site. That’s simply too much.

Ah you’re right. Seems like BitMEX is still too small for large traders. In that case, I recommend including the expected cost of clawbacks in your analysis. I think it’s a great idea to manage risk with derivatives, but I’m not sure if sufficiently reliable derivative exchanges are available for cryptocurrency at this point in time.

Not sure if you’re referring to my proposal to ensure safe and reliable NBT exchange for cryptocurrency.

The basic idea is to automatically execute a mirrored trade on a trusted exchange before finalizing a NBT/Cryptocurrency trade on the “Nu” exchange. Instead of having to actually hold large quantities of various cryptocurrencies that custodians might not trust, we would only purchase those cryptocurrencies when necessary to complete a transaction. This plan would have a fairly significant upfront cost (particularly in terms of time and effort to create the exchange), but would ultimately be less expensive (and potentially profitable) in the long run.

thanks @GreatLogic. But no it was not that post. In the post I remember, to trade between NBT/crypto, the exchange will need to hold down funds on two pairs and execute them in an atomic fashion so there won’t be price slip. It would be like direct trading between NBT-crypto without prolonged price exposure.

While I understand your critique of converting crypto currency funds to NSR (I was marking out that problem in the “identified risks” section of my post), I don’t like the dependency on a party outside of Nu that is being created by involving Okcoin.
Short-term this seems to be the better, the feasible way.

But I think a discussion about the “convert to NSR approach” could be worthwhile when looking for a long-term solution in a world with high NSR trading volume.
I just realized that custodial grants of NSR are being discussed (although my approach differs in some aspects):

If I screw up this thread by continuing the discussion here, I create a new one or ask for my post(s) to be shifted to a new one. My proposal might be a possible “solution to custodian loss of value due to BTC exposure”, but of a different kind.

The longer I think about that “convert to NSR and burn the NSR” approach, the better I like it - did I oversee obvious problems (beyond the dependancy on the trading volume?
Burning incorporates the value of the alternate currencies (even converting parts of USD to reduce the counterparty risk could be considered depending on the market situation) into the Nu network.
It would be required that burning the NSR automatically creates an address.
To this address the formerly burned NSR are spawned by the passing of a special custodian grant.
That grant would need a significantly lower threshold to be passed to allow quick reaction to a need for the NSR.
The grant for that needs to be created immediately after the burning to be prepared for the need of the NSR.

This grant can be put in feeds and distributed to clients which subscribed to them without share holder attention and hence very quickly.
Although it might appear not very secure to have a grant with a low pass threshold, I think even a very low threshold to pass this grant is more secure than keeping the currencies at an exchange or in the wallet of a custodian.
Even allowing such a grant to pass after 2 blocks in a rolling window of 30 blocks is better than not burning the NSR.
And a low threshold allows quick emission of the NSR. Remember that this is for tier 3 of my proposed solution, there are funds in tier 1 and tier 2 as well!

In case such a grant passes “accidently” because a feed includes the hash without need, there’s nothing more to do than to burn the NSR once more.

In case a custodian disappears for whatever reason, the value that is otherwise kept to the exchange account and maybe wallets (and maybe lost forever!), is still in the Nu network.
In that case the automatically created grant needs to be withdrawn and a standard grant to create the NSR at a different address with the standard passing threshold needs to be created.

Problems I’m aware of:

  • the very low passing threshold seems could be usable for exploits. It’s required to tie such an address to the NSR burn transaction (sounds like a protocol change), as only this poses no additional risk; the automatically generated address is in control of the custodian where the funds were without burning anyway. In case the custodian disappears, a standard custodian grant with standard thresholds would be used to create the NSR.
  • to recreate the burned NSR quickly (faster than 50% of 10,000 blocks; minimum ~3.5 days) the protocol needs to be changed
  • feeds are not yet implemented, but as this solution (if it turns out to be a solution) is for the future, that doesn’t hurt
  • abuse of the automatically created address is possible; scenario: custodian disappears, new custodian grant is created to issue NSR, evil actor creates a custodian grant with the “old” address and the “special grant low threshold”. NSR are created and lost to the old address (if the address is in control of the attacker it’s even worse; this would incentivize “playing” the role of custodian to abuse that at a later point of time).
    Only if it’s possible to deactivate the “old” adress by tying a “new” address to the burn transaction (effectively replacing the “old” address) this attack vector is mitigated

This is a lot of talk. Forgive me wasting your time if it’s complete nonsense.

I think the problems of currency burning and holding reserves are linked.

In a 0% reserve scenario, the value of NuShares is ostensibly in the forward revenue stream Nu can generate, that are eventually to be paid out as dividends. At the time of dividend issuance, the value of NuShares should decrease by the amount of dividend distributed.

In the case where there are too many NuBits, this implies no more making of NuBits, therefore there is no revenue coming in, and the value of NuShares will drift to zero. (Correct my logic if it is wrong.)

How then can we sell NuShares to mop up NuBits, when the market would already be pricing in NuShares of value zero? It would have to be based on 100% speculation and faith that the Nu ship would be righted.

If there are reserves and/or other revenue streams, at least there is some NuShare value to dilute.

Granted a large part of this is that Nu was originally to be exchanged with USD, yet we are chasing cryptos. However this may be a good battle test of the solution if/when/as future pegged currencies have issues, and we are at a million times current scale.