Feedback loops and Analysis of Bitshares architecture in relation to Smart-Assets sustainability
Eliminating the negative impact of Global Settlement events (GS) is certainly a priority for the Bitshares ecosystem. Currently, GS is widely regarded by the Bitshares community as a natural consequence of harsh market conditions, bad debtors and poor marketing strategy. Contrary to this perspective, I maintain here, that GS might be the consequence of some other more fundamental factors in the dynamics of the whole Bitshares protocol. I postulate that if these factors are modified, the probability of a GS or a similar event is drastically reduced almost immediately and may even tend to disappear in the long term. These factors include, but may not be limited to, two feedback loops currently present in the dynamics of workers’ financing and Margin calls. This work attempts to elucidate the architectural nature of the problem, out of which GS is only an inevitable result. Also, a new mechanism is proposed.
Statement of the problem
Feedback loops appear everywhere, especially in social and biological networks and their effects in a given system are usually dramatic. One very important aspect of them is that for a given system with a given set of feedback loops, its behavior can be modeled by a set of differential equations, and thus, although chaotic at first sight, its fairly easy to predict its long term equilibrium state. Thence, the importance of having them right in any complex system design. If the reader is not familiar with feedback loops’ nature and effects, he can visit Wikipedia.
In the case of Bitshares, the whole network is exhibiting a feedback loop that pushes BTS price down by the current margin call mechanism. In fact, smart-assets such as BitUSD and BitCNY can be borrowed into existence by anyone with a Bitshares wallet. In order to guarantee the value of those smart-assets, Bitshares protocol requires BTS to be left in deposit from the borrower as collateral. This frozen collateral plus the debt obligation constitute a margin position. When BTS price goes down, some margin positions will inevitably become poorly or insufficiently collateralized.
Here comes the crucial part. Bitshares protocol removes those bad-debt positions by selling their BTS into the market at a premium price. This premium price is enforced to guarantee the sale and is frequently overcome by trading bots running through the DEX, constantly. Competition between bots and margin calls brings BTS price even lower, triggering at its turn more margin calls, more BTS dumping, and so on and so forth, etc. I will refer to this feedback as “Margin call Feedback Loop (MFL)”. MFL is, of course, much stronger and damaging during bear markets.
A second, harmful, negative-feedback loop on BTS price comes via Bitshares’ workers. Worker is the name given by the community to the projects, financed by the blockchain itself, that contribute to the improvement of the ecosystem. Paradoxically, the more workers are financed, the more BTS is sold, on a daily basis, by those very workers in order to obtain BTC or fiat to pay for their bills. So, the workers’ need to sell big chunks of BTS daily is driving the price down in the immediate term, and this is mostly independent of market conditions. In a bull market this effect is somehow lessened by the stimulus of the people behind the workers to keep some BTS and profit from the rise in price. However that’s not always possible due their own immediate financial obligations. Under bearish conditions, on the other hand, the effect from this constant ‘dump’ of BTS into the market may be dramatic. For the purpose of this writing, I will call this second feedback “Worker Feedback Loop (WFL)”.
The real impact of the these two feedback loops can only be grasped if one considers that during bear markets they amplify each other. Indeed, a sole big worker can generate a big BTS sale that brings the price low enough to activate several margin calls that, at their turn, will force the protocol to offer BTS at a discount. This new lower price frustrates the worker’s initial sale and thus, motivates it to bring its price even lower. A downward spiral in price is so generated. It is precisely this spiral what should be the focus of any serious attempt to deal with global settlements in all native Bitshares smart-assets. Importantly, and somehow fortunately, this self-driving-downward effect is a direct consequence of the Bitshares architecture and not totally dependent on additional factors such as Marketing, nor dependent on out of control factors such as overall market conditions or bad debtors.
If MFL effects are to be minimized then, to the greatest extent possible, BTS from margin called positions must not be immediately dumped into the market, period. Any other alternative in which that BTS is right away offered for sale will end up in a MFL in some way or the other. Because weak positions still need to be liquidated or recollateralized, and holders of those positions may not be willing/able to perform any of those actions, implied is the need of new method for bad-debt management.
Some members of the community, including myself, have suggested the implementation of a Settlement Protection Fund that takes over weak debts. Once the fund takes over those positions, it must not liquidate them for, otherwise, it would make no difference to the current protocol and would not eliminate the incorrect feedback. Then, a second implication is that the Protection Fund must be constantly replenished in order to recollateralize the debt positions it acquires.
To me, that’s a first solid result of the present analysis. A Settlement Protection Fund is not an alternative, it’s a necessity if we are to get rid of MFL. So, I propose that whenever a position with a CR < 1.10 or some other critical CR appears, something like the following is implemented:
First, the Settlement fund takes over the bad position, both the debt and the collateral. The choice of words here is important. The fund doesn’t buy the position, it takes over it. The former holder doesn’t own the position any more, nor he has any debt to pay. Also, the position doesn’t disappear, it still exist with a new owner: the fund. Because the fund doesn’t buy the position, it needs no initial funding. Also, because the fund is taking over positions whose CR is around 1.10, the collateral of the fund would tend to be at least that value. I say ‘at least’ because given that the fund is receiving a fraction of the network fees, this collateral should tend to increase, except in cases of harsh decreases in BTS price. Of course, the big advantage of having the fund taking over instead of dumping the position in the market is eliminating the MFL cited above.
So, the settlement fund would constantly take over bad debt and would hold itself a margin position that would be equivalent to all the BTS and all the debts it holds at any given moment. Also because it would constantly replenish its collateral, one could consider the fund a responsible and capable debtor.
What happens if BTS price increases dramatically? In this case the fund’s CR will also increase. I propose that whenever this CR tops a given value, let’s say 3, the excess of collateral is used to buy some smart-assets in the market, assets like BitUSD, BitBTC, etc. These new bought assets can then be put in a second fund, whose mission should be to pay for workers, not in BTS as today happens, but directly in stable coins. This last step should alleviate the effect from the other feedback loop: the WFL.
One could argue that it serves no purpose to pay workers in, for example, BitUSD, because those workers would have to then sell that BitUSD for BTS and then sell that BTS for BTC, dumping BTS as it happens today. This is however not a sound reasoning from my perspective because under this new protocol a two-steps process is involved. First step involves buying BTS and the second step selling BTS which is very different from the current protocol in which BTS is only dumped. Also, mechanisms for directly switching from BitUSD to fiat or BTC are currently being developed. Overall, I think that payments of workers in smart-assets such as BitUSD, BitCNY or even BitBTC should be mandatory whenever possible, not optional. This way BTS price would only be minimally affected daily by WFL and would bring the added benefit that, if workers want to hold BTS, they would have to buy it, pushing BTS price upwards .
On the other side, what happens if BTS price declines sharply? In that case, given that the fund would constantly take over the worst collateralized positions and it would do so for as long as it has a CR>1, a pretty extreme scenario occurs when fees that replenish the fund are insufficient and the fund’s collateral falls below 1. In that case, instead of only taking over all the positions whose collateral falls below 1.10, the fund could begin taking over the next low collateralized positions that fall below MCR. That is, those whose collateral falls between 1.10 and 1.60 in the current BitUSD model. This second-level take over doesn’t have to share the characteristics of Partial Global Settlement, in which all the positions below MCR are simultaneously taken, instead, it could take positions one by one, only until the fund’s CR recovers above 1. Because by that moment all the worst debt positions would have been taken and the fund would still be constantly replenished by fees, there would be no need to take over more positions than what is necessary to bring back fund’s collateral to 1. Also, if the former owner of a taken-over position wishes so, he could get his position back from the fund by adding some collateral and paying a small fee. Some kind of warning could be implemented for this kind of debtor, under this situation.
The worst case scenario occurs when a fall in BTS price is so precipitous that all positions with a collateral below MCR get taken and yet the fund’s CR falls below 1. Notice that this is a very extreme and unlikely scenario which, fortunately, is not as bad as current management of GS. Because, in that worst case, although fund’s collateral would fall below 1, all positions with collateral below MCR would have already been taken care of. So, the only bad debtor would be the fund itself. In that case, having the fund no other bad position to take over, all that would have to be done is wait until the inflow of fees replenish back fund’s CR; and, this would not necessarily affect the peg because for the only bad debtor in the whole ecosystem (the fund), there would be many other positions with collateral above MCR. So, in average, the peg could be maintained. Of course, during these very extreme conditions in which the fund’s CR falls below 1, forced-settlements would have to be temporarily suspended, for it wouldn’t be possible to deliver them without affecting healthy collateral positions. As a consequence, anyone willing to get BTS for a Bit-Asset, would have to do it at market value, helping in this way to bring BTS price upward.
Potential benefits of the new protocol
The above presented protocol has the following potential benefits, even under very extreme market conditions.
- Manages weak debts without amplifying volatility and without the need for GS.
- Eliminates need for GS protection and the corresponding cheating in price feed from either the witnesses or the protocol itself.
- Eliminates the Margin Call feedback loop that currently has the BTS ecosystem inflicting self-damage.
- If a second fund is created in order to accumulate smart-assets to pay for workers, this new protocol would reduce the also harming feedback loop from Workers selling their pay.
- This new protocol protects the Native smart-assets peg.
- Except under extremely harsh market conditions, all of the above benefits are obtained without affecting current normal forced-settlement protocol.
- This new protocol has been written thinking of the users of the DEX as both its clients and shareholders. So, they must be treated accordingly. Consequently, no additional fees of penalties are mandatory for those who fell in a bad position.
- If successful, this new protocol could help the revival of long ago inactive assets such as BitEur and BitBTC.
Most probable consequences of not modifying the current protocol
As stated before, feedback loops are ubiquitous in complex systems. Complexity and feedback come hand in hand. Also, feedback loops can be easily modeled through math and its long term behavior can be easily predicted and calculated given sufficient parameters. Because the Bitshares ecosystem has been running for a long time, even without need for any further mathematical model, one can get some insight of that long term behavior by looking at the price evolution of BTS. A negative feedback on the price by MFL coupled with another negative-feedback by WFL, will keep dwindling the bull trends and aggravating the bear trends. Eventually, there would be no stimulus nor enough investors to borrow smart assets into existence. Empirical evidence of this might be gathered from the BitEUR and BitBTC cases. Finally, failing to keep on with smart-assets would be a fatal wound to the Bitshares ecosystem.
Yet another weakness in the current protocol
Some very serious players in the Crypto-Space would prefer CEXs to disappear. Many new people comes to Bitshares precisely looking for an alternative to CEXs and, since its very inception, Bitshares could have been a serious competitive threat to big CEXs such as Binance and Poloniex. I am afraid that won’t happen anytime soon because, ironically, under the current protocol these companies hold enormous power on the DEX, both because of their voting power and because the price feed is totally dependent on them. If at any given time, a smart asset such as BitUSD is prone to be globally settled, or to lose its peg. All these big companies have to do is short a small fraction of their big holdings of BTS in order to wreak havoc on the community. Current Bitshares protocol is designed to depend on its competitors. Furthermore, the bigger the competitor, the strongest our dependence is on him. That is not a good strategy.
That problem, however, may be the topic of another future analysis.
From the Mountains of South America, I wish you all the best...
The author declares no interest of any kind whatsoever from the Bitshares community from the above analysis. If you are a member of the Core Bitshares team and you aprove these ideas, feel free to create or incorporate them into the corresponding BSIP.
Donations are dearly welcomed