Tier 4 Fund Management Discussion

mhps 2015-10-18 14:42:59 UTC #21

The problem with Tier 4’s dealing with Tier 1-3 is that T4 fund is from a sigle entity and T1-3 are managed by many LPs. How can T4 fund be injected to T1-3 in a timely and responsively, and to where it is needed? Will T4 managers have to be all-seeing-all-knowing to do a good job? That seems impossible?

The bottom of the problem is that current T4 liquidity is centralized. To solve the problem I have an idea of a different approach.

What about using T4 fund only as reward/interest that LPs get? T4 doesn’t provide liquidity fund per se, it provides incentive to make T1-3 materialize. For example T4 fund pays for every pool’s interest, every minute via a t4bot, according to global buy and sell side liquidity. T4 fund will deal with limited number of pools’ interest wallet, instead of a huge and unknown number of individual LPs’ liquidity accounts. To be safe, a t4bot only holds a week’s worth of interest. T4 managers will replenish t4bot’s wallet every week from the multisig address, and ask for fundings from shareholders for T4 operations per 3 months.

There are some technical improvements for example there can be several instances of t4bot controlled by different individuals for redundancy and decentralization.

Nagalim 2015-10-18 14:50:19 UTC #22

I don’t see how asking tier4 to keep account of every LP is simpler than making deals to rebalance nulagoon (tier 3) and buying nbt on-exchange to burn (tier 2 and 3).

mhps 2015-10-18 14:59:04 UTC #23

Are addressing to my post? If yes I didn’t ask T4 to keep account of every LP (liquidity provider) but every pool (operator’s) interest wallet.

Nulagoon is just an example of every MLP. What about ALPs? My idea is dealing with every pool, uniformly and automatically.

Buying NBT to burn is the job of T6, no?

Nagalim 2015-10-18 16:20:35 UTC #24

NuLagoon’s funds are separate from NuLagoon’s monthly cost to Nu. A rebalancing for NuLagoon has to do with their funds (buy vrs sell side), not their monthly cost. In your mechanisms, how would Tier 4 deal with a broken peg? Pay more to LPs to provide funds? That’s just the parking mechanism to an extreme.

T4 is originally designed as the developer fund, and it contains funds held in units of account external to the Nu system. It is a reserve and should be mobilized during periods of extreme network stress to buyback and burn NBT. T6 is indeed shouldered with the responsibility of burning NBT during periods of network stress, but cannot possibly be used in extreme cases without triggering a black swan event. That is why T4 should be mobilized to rebalance NuLagoon (T3) and buy back nbt from T1 custodians directly. In my opinion, the T3 rebalance should have higher priority than the direct NBT buyback, but the absolute fastest way of mobilizing funds in a peg brake is direct insertion into T1.

For example, someone slowly buys 150,000 NBT from the Nu network over the course of the year. Then, they put up every NBT they have as a sell wall all at once at $0.5 on NBT/BTC. Is Nu really going to sell $150,000 of NSR all at once, or are we instead just going to pump all developer funds at that wall? Note that if we decide to pump developer funds at that wall we could recover the peg near-instantly and still have money left over in T4.

masterOfDisaster 2015-10-18 17:18:23 UTC #25

No matter how tier 4 was designed - Nu needs a buffer to buy back NBT; tier 4 is in the ideal position for that.

This is my opinion as well.
But in the end it doesn’t matter much, as the market can move funds between the tiers 1 to 3 quite fast.
The more urgent NBT need to be removed from the market, the lower tier to place the BTC can be chosen.

I’d be happy to buy as many NBT at $0.5 as I can afford as long as I can hope to sell them for $1 to Nu.
Nu could happily do the same.

At $0.5 only $75,000 in NSR would be required to buy the wall.
But I don’t think that the market depth of NSR is deep enough to dump $75,000 in NSR on the market without killing the NSR price.

In my opinion that’s exactly what tier 4 is for: buy NBT if the sell side is much bigger than the buy side.
Nu needs to define at which ratio action is expected from tier 4.

Tier 6 is what either

refills tier 4 (NSR get sold for BTC, the received BTC fill tier 4 buy side) or
interacts with lower tiers through seeded auctions (NSR get sold for NBT, NBT get burned)

at least in my opinion.

Nagalim 2015-10-18 17:24:07 UTC #26

You’re right about it only being $75,000 NSR sold, but that’s still way more than we can efficiently do in a short term without black swanning.

I disagree. Tier 4 is for developer funds. The only time Tier 4 should be used to support the peg is if there is actively a peg break scenario and Nu stands to lose its reputation. Tier 6 should be used to rebalance the global peg.

Refilling tier 4 should come from selling NBT in my opinion, not selling NSR. It should occur when Nu is expanding, either in the economic or the development sense. The genesis refill was of course the IPO, which we are performing buybacks with currently.

masterOfDisaster 2015-10-18 17:56:30 UTC #27

Why should Nu keep developer funds in a volatile form?
I see demand for a tier 4 and a tier 4 sell side that interact with lower and higher tiers.

Nagalim 2015-10-18 18:00:07 UTC #28

That’s tier 3 by definition though. Tier 4 is developer funds that also happen to be a reserve. It’s just that most developers in the crypto world prefer to be paid in BTC.

Interpretation of the liquidity tiers, a waterfall model, triggers, metrics and actions

masterOfDisaster 2015-10-18 18:21:43 UTC #29

Nu does not by definition hold tier 3 funds.
BTC could be exchanged close to payment with granted NSR.

edit: sorry for this extremely short reply. It was not meant unfriendly. I was typing on mobile phone. I think I’ll start a new thread that tries to supplement this discussion here by trying to draw a bigger picture in which the T4 fund management discussion hopefully fits.

mhps 2015-10-19 04:33:45 UTC #30

So? Does it not solve the problem?

Exactly. If the price is not going t bounce back to $1, I’d rather keep the developers paid than extending the peg for another few hours/days, because the developers are the source of hope.

We may first need to agree on this first: when is the T4 fund, which is mainly for paying developers, to be used as liquidity support fund? T4 managers need to clearly understand this to do their job.
To me these two things are an odd couple.

masterOfDisaster 2015-10-20 17:35:26 UTC #31

Do we have a chance to speed the development of tier 4 fund management up?
I’m asking because of this:

Nagalim 2015-10-20 18:25:11 UTC #32

Can we have a thread like this one but for Tier 6? I think we might be able to come to consensus for Tier 6 more quickly than Tier 4.

masterOfDisaster 2015-10-20 18:39:32 UTC #33

I don’t see the reason or need for that.

Tier 6 consists of NBT or NSR that are granted by NSR holders for a reason.
The addresses where the NBT and NSR are granted could very well be tier 4 multi signature fund management addresses.
Conceptual the 40 million NSR that will be burned soon would fit in tier 4 buy side - if the tier 4 buy side would hold NSR next to BTC.

At least this is my interpretation of the liquidity model.
I see no reason for having a(n additional) tier 6 management team - effectively the NSR holders are the tier 6 management team!

Nagalim 2015-10-20 18:45:59 UTC #34

Tier 4 holds BTC currently and is reserved for paying developers. Tier 6 holds NSR with the sole purpose of defending the peg and is about to be burned. All I’m saying is that we need a multisig team to hold a reserve of NSR with the sole purpose of defending the peg. If you don’t want to call it T6, fine, let’s call it the Magical Unicorn Signers, whatever, let’s just get a multisig group with control over an NSR address with the sole purpose of supporting the peg according to some simple rules (like those that JL just posted).

masterOfDisaster 2015-10-20 19:18:18 UTC #35

Reserved for that, but not limited to it.

I beg to differ. Tier 6 creates NBT or NSR if necessary - by grants - to defend the peg.
Using feedback loops from lower tiers reduces the risk for the peg.

I agree that an NSR reserve with the sole purpose of defending the peg is useful.
I recognize tier 4 as the ideal layer for that; multi sig included.
I like “Magical Unicorn Signers” and would like to call the tier 4 fund management team that.
What about the draft of rules I proposed?

Nagalim 2015-10-20 19:43:37 UTC #36

Tier 6 doesn’t print nsr or nbt, shareholders do. The tiers are each a separated reserve. Tier 6 is an nsr reserve held for the sole purpose of supporting the peg in times of need.

Sorry, if you summarized the rules into a simple list somewhere I must have missed it.

masterOfDisaster 2015-10-20 20:36:09 UTC #37

It is my understanding as well that tier 6 is dealing with NBT or NSR printed by shareholders.
But I don’t expect NBT or NSR to be kept by custodians on tier 6 - for what reason?
They are granted on demand.

I wouldn’t call it summarized rules, but it’s here:

Trying to summarize it:
Tiers 1 to 3 are outside of direct control by Nu.
Tiers 4 to 6 are in full direct control by Nu.

Tiers could be interpreted in a waterfall model (this is my interpretation, please add to it if you like; the triggered actions are just a draft, a proposal).
Each tier interacts with the adjacent tiers; tier 5 is something different.
Nu provides an incentive for putting funds on tier 1.
Tier 1 drains funds from tier 2, which drains funds from tier 3, etc.

Tier 4 holds funds for paying development operational costs and keeping the peg (BTC and NSR to reduce the dependency on BTC; ratio between BTC and NSR needs to be defined - why not 50/50?).
If the ratio of buy side sell side (or vice versa) is below 1.5 nothing happens.

Connection of tier 4 to lower tiers:
Between a ratio of 1.5 and 2 seeded auctions try to balance the sides (most effective likely on tier tier 3, less effective on tier 2, least effective on tier 1). Funds are taken from tier 4 (buy or sell side/ NSR or NBT). The seeded side is the smaller one.

Above a ratio of 2 funds from tier 4 are being injected to tier 1 to support the peg (BTC or NBT; depending on the side that needs support).

Connection of tier 4 to higher tiers:
If tier 4 (buy side) is below 12% of the value of NBT in the wild, shareholders are expected to fill it. A corridor of 3 percent points tries to limit the involvement of NSR holders.
Once tier 4 buy side is below 9% NBT value in the wild tier 4 buy side (multi signature fund management address) gets filled with an NSR grant (tier 6 triggered).
Once tier 4 buy side is above 15% of NBT value in the wild, NBT get burned by the fund managers.
The same rules could apply to tier 4 sell side (then tier 4 buy/sell side should try to get balanced before NBT get burned or NSR granted), but as funds on tier 4 sell side pose no default or volatility risk they can be vastly different.

On tier 5 shareholders are expected to raise parking interest above 0% if the average sum of tier 1 to tier 3 in the last time frame x of sell side liquidity is bigger than buy side.
To go into greater detail about parking rates is out of the scope of this attempt to summarize the rules - I don’t even have a detailed proposal in the verbose version.

Nagalim 2015-10-20 21:05:45 UTC #38

Do you have a rigorous definition for this? Remember that liquidity data from tiers 1-3 can be manipulated fairly easily (just get shareholders to pass a 1 NBT grant).

My instinct is to call this false. The only thing that shareholders are directly in control over is the creation of NBT and NSR, which leads me to a different model than the waterfall model:

Nu can create NBT or NSR at will to incentivize reserves with different intents and rules.

T1 is the front line reserves. Highly decentralized
T2 is transient backup reserves for the front lines. Decentralized.
T3 is cold reserves for the front lines. Mildly decentralized.
T4 is cold reserves for developers. Can be mobilized for the front lines. Multisig.
T5 is negative nbt reserves that can be considered out of circulation. Highly decentralized.
T6 is operational cold reserves. Can be mobilized for the front lines. Multisig.

Motions and grants affect these tiers directly and can cause funds to flow between them.

The operational cold reserves are for converting NSR to buy liquidity (or directly dealing with NBT:NSR via seeded auctions eventually) in a short period of time. Effectively, our % reserve is:
[T1+T2+T3+T4+T6]buy / (NBT marketcap - [T1+T2+T3+T4+T6]sell - T5)

masterOfDisaster 2015-10-21 08:09:08 UTC #39

With “NBT in the wild” I mean the amount of NBT sold from Nu to people:

nud -unit=B getinfo | grep supply
    "moneysupply" : 4625463.0383,

The 4,000,000 NBT of the FSRT needs to be subtracted from that number, because obviously the FSRT is still in possession of Nu. The 40,000 NBT management fee for the FSRT are a part of NBT in the wild. They will not need to get sold to enter the market, or the tier 1-3 sell side, respectively.

As a result I consider 625463.0383 NBT in the wild.
This number can’t be manipulated except for burns and grants.

Maybe I should have written funds on tier 4 to 6 are “(temporarily) owned/held” by Nu.
I’m not saying that Nu possesses all the funds (e.g. tier 5). But on tier 4 and 6 Nu possesses the funds is free to do as NSR holders decide.
Funds on tier 1 to 3 are “managed” by people who possess NBT and BTC.

I don’t see why T2 is less decentralized than T3, because all funds on tier 1 to 3 are in the hands of the same people - more or less.
The NBT in these tiers have been sold from Nu.
The BTC in these tiers is provided by LPs, because they are incentivized (by Nu) to do so.

T4 is reserves for developers, but for ongoing costs and supporting the peg as well.

T6 is NBT or NSR granted by NSR holders.
Why have another fund management team dealing with multisig addresses, if the NBT or NSR can be granted to tier 4 multisig addresses as well? That makes things more complicated, but what’s the benefit?

mhps 2015-10-21 11:05:09 UTC #40

Your model does not seem in contradiction to @masterOfDisaster’s significantly to mee.