Greetings everyone!
Long ago, when I found out about Bitcoin, I was very excited about the concept of a new generation of money powered by a decentralized ledger technology (just like everyone of us was, I suppose). The time has shown, however, that Bitcoin has failed to become a monetary asset and turned into a hedging tool or a speculative asset instead.
In 2018, after completely realizing that Bitcoin wouldn’t substitute fiat money, I began working on a cryptocurrency that would succeed in reaching that goal. I had to develop a completely novel consensus protocol from scratch, which was an extremely challenging task, but the most challenging issue was the development of an appropriate economic model.
Meanwhile, some developers of algorithmic stablecoins tried to solve the same issue with quite underwhelming results (Hello, UST!). The mere idea of stablecoins is still inappropriate for an asset that can become a new generation of money and substitute fiat because a derivative of an asset cannot become a full-scale alternative to the said asset, not speaking of depegging risks and the vulnerability to speculative attacks.
After all, to create the money of the future, we need to completely detach decentralized currencies from their fiat counterparts by upgrading the concept of a stablecoin to the concept of a stable economy, the difference being wider capabilities of the algorithm: instead of simply reacting to shifts in demand against a particular asset to maintain a preset exchange rate, it should be capable to counter shifts in demand coming from all possible channels and directions and maintain targeted values of the variables used in conventional macroeconomic policies, namely inflation and/or output growth.
Simply put, to create a decentralized money we need to create a decentralized Fed or, as I call it, a Decentralized Algorithmic Central Bank (DACB). Apparently, we cannot adapt the Fed’s discrete monetary policy since discretion is incompatible with decentralization, which is why DACB is inherently a model with an incorporated feedback policy rule.
Herein, I present the economic model of a Decentralized Algorithmic Central Bank, which allows targeting inflation via a feedback policy rule similar to the Taylor and McCallum rules. Unlike stablecoins, which are only capable of maintaining the exchange rate against a particular asset, the model performs as a fully-fledged central bank and is a generation ahead of all stablecoin concepts. Besides, the model doesn't fall under the stablecoin definition within MICA regulations and operates with native utility coins of the system, so the security regulations are also circumvented, which is a huge advantage over any existing stablecoin.
DACB VS Fed
Simply put, we can describe the Fed’s discrete policy as targeting inflation and/or employment via the control of the short-term interest rate, which serves as an intermediary target. The Fed uses four main policy tools to keep the interest rate within the targeted boundaries: the interest on reserves and ON RRP rate to set the lower boundary, the discount rate to set the upper boundary, and open market operations to adjust the available reserves in the banking system and keep them sufficient.
My DACB implementation directly targets the short-term nominal interest rate, which also allows to indirectly target other economic variables that are correlated with the interest rate, the primary one being inflation via the Fisher effect. To control the target, the algorithm uses the adjustment of the monetary base and velocity of circulation via the policy tools.
The issuance rate is the main policy tool. By increasing it, the algorithm performs the expansionary measures pushing the targeted value upward. By decreasing it, the algorithm executes stage one contractionary policy pushing the targeted value downward.
The contractionary rate is the secondary policy tool. It is activated during stage two contraction, when the issuance rate reaches its lower boundary. Stage two contraction reduces the velocity of circulation by increasing the stake pool, which pushes the targeted value downward.
Features of DACB Model
DACB Incorporates Feedback Policy Rule
DACB is a generation ahead of stablecoins, for it can not only follow the exogenous course by being pegged to allegedly stable off-chain assets but sustain a targeted equilibrium endogenously.
DACB incorporates a variation of a feedback policy rule with a targeted interest rate similar to the Taylor or McCallum rules to create a functional stable economy that satisfies the needs of all participants. With such a functionality, the model can make Dynemix the money of the future.
DACB Is Independent and Decentralized
All stablecoin designs feature a common downside: they always rely on external sources and are pegged to exogenous assets.
The DACB model makes a major breakthrough by introducing a concept that relies exclusively on on-chain data, is fully algorithmic, and is not related to any particular off-chain asset (USD, gold, SDR etc.) Instead, the model operates in conjunction with the entire global economy.
DACB Works with Native Coins
Algorithmic seigniorage-style stablecoins are paired with special assets to balance the supply against the demand.
The design of such assets endows them with the properties of securities, which puts the system under the control of regulators, creates significant legal obstructions, and severely undermines decentralization.
DACB operates directly with native utility coins, which allows to avoid all the respective difficulties occurring with the introduction of ancillary assets, including sanctions from financial regulators. For example, it’s not classified as a stablecoin (neither ART nor EMT) under the MICA regulation or as a security token under MiFID II.
DACB Is Highly Egalitarian
With the help of helicopter coins, the model impedes the concentration of power and doesn’t allow strong actors to sustain their influence endlessly simply by staking all their coins and increasing the relative share in possession.
Since banks and credit are removed from the model, strong financial actors can no longer take direct advantage of a monetary expansion and receive or create money out of thin air.
Helicopter coins also stimulate the demand from the consumer side, which allows to circumvent many credit-related monetary issues.
DACB Can Be Potentially More Efficient Than Fed
The effectiveness of a monetary policy is fundamentally limited by the credibility of a monetary authority. In practice, credibility can be a weak point of a monetary policy. Despite being commonly positioned as independent actors, CBs often suffer from governmental interference, which can undermine credibility and render applied measures less effective. The governing boards of CBs can also be affiliated with commercial banks, which further aggravates the issue.
The DACB model is deprived of that flaw for it is completely decentralized and trustless. There can be no question of credibility or impartiality of an algorithm that is incorporated into a blockchain system, which makes the monetary policy highly efficient in terms of managing expectations and driving the behavior of participants in a desired direction.
DACB Can Seamlessly Operate with Conventional and Islamic Banking
Most of the countries that follow Islamic financial doctrine suffer from severe economic difficulties for Islamic banking is not consistent with conventional banking, which is based on credit.
Since my model doesn’t rely on credit, it is fully compliant with the Islamic principles and can be adopted on the full scale by Muslim countries, thus solving the problem of seamless integration of the Islamic banking into the global financial infrastructure and allowing to overcome economic difficulties that many of Muslim countries are currently facing.
Technical description
Here's some boring technical crap that nobody will probably read, but I'll still leave it here just in case. If you are into it, though, you can go straight to the full paper.
DACB, Dynemix Blockchain and Liberdyne Messenger
Dynemix is a decentralized, permissionless, account-based blockchain system powered by a unique Proof-of-Stake BFT consensus protocol. Dynemix is the first DLT system to incorporate a DACB model. The DACB model of my design requires a number of specific technical features from the underlying blockchain, which is why it cannot be implemented into any existing platform, and which is why I do not give it a separate name and simply call it “the economic model of Dynemix”.
Dynemix features a low-level transport protocol for the delivery of messages to offline users called DOG. The Liberdyne messenger incorporates the said protocol in its design, and thus provides a practical solution for the distribution of helicopter coins, which are a crucial element of my DACB’s design.
Helicopter Coins
Current gen blockchains (and other DLT designs) feature a common coin distribution model: newly mined (minted) coins are distributed among block producers as an incentive to process state transition and keep the system secure.
Dynemix introduces a novel approach: only a share of newly issued coins goes to the minters of the block, whereas the rest are dispersed among users of the platform, thus becoming an embodiment of the concept of helicopter money introduced by Milton Friedman.
The share of coins dispatched to minters called a minters’ share can be arbitrarily adjusted by the algorithm within the predefined boundaries, which serves as one of the tools of the monetary policy, as it, together with the issuance rate, defines the nominal interest for staking.
Tools of Monetary Policy
The algorithm conducts the monetary policy via the following tools:
Issuance rate (ki) – the annual rate at which the coin supply grows.
Minters’ share (ks) – the share of newly issued coins that is dispatched to minters as a reward for a block creation.
Contractionary rate (kc) – the annual rate at which the targeted size of the stake pool (S*) is being increased.
Policy Rule
The algorithm conducts the monetary policy according to the following general rule:
ki+kc = 2(I*-I)+ki*
Where:
ki is the issuance rate;
kc is the contractionary rate;
I is the current nominal interest rate;
I* is the targeted nominal interest rate;
ki* is an assumed issuance rate that maintains the targeted inflation rate and potential GDP.
Suppose we agreed that the potential GDP growth rate is 3%, the targeted inflation rate is 2% and the natural real interest rate is 2%. Then we can rewrite the equation as follows:
ki+kc = 2(4-I)+5
Default Equilibrium
If we target economic variables to reach the values stated above, we should set the algorithm as follows: the issuance rate should be set to 5% to provide 2% inflation assuming 3% economic growth; to target 4% nominal interest (which equals to 2% real interest under 2% inflation) and 20% stake pool size (which provides enough security), the minters’ share should be set to 16%.
Under the stated setup, the interest rate curve would look as follows:
We can see that with the growth of the stake pool (S), the interest rate for stake holders (minters) declines. As more coins are locked as stakes, it becomes less profitable to participate in minting, and some of stakeholders will unstake their coins to invest them into higher yielding assets. There is a point of equilibrium at which S holds at any given time. This point represents the current interest rate in the economy. On the chart above the interest rate is 2% in real terms or 4% in nominal terms, which is the targeted r* for the algorithm.
A change in the macroeconomic setup will cause a respective change in the desired interest rate, and the size of the stake pool will shift either upward, which will trigger the expansionary policy, or downward, which will trigger the contractionary policy.
Expansionary Policy
Suppose the inflation rate is expected to drop from 2% to 1%. According to the Fisher effect, I will drop to 3%, which will push the equilibrium to S=26.6. Under I<I*, kc=0, and the algorithm will recalculate ki according to the rule above: ki=2(4-3)+5=7. ks will be also adjusted to offset changes in S*. Therefore, for 1% expected drop in inflation the algorithm increases the issuance rate by 2%, which provides a counterpressure on the interest rate and pushes the system back to the optimal equilibrium. The coefficient of the rule (a) can be arbitrarily adjusted. I suggest a=2 for the general rule as a more aggressive reaction than proposed by Taylor (1993), namely a=1.5, due to the specifics of the blockchain setting.
The same reaction will be applied to the pressure on the interest rate coming from other channels. For example, if the exchange rate of our coins against exogenous assets (say USD) is expected to shift upwards, our currency becomes more attractive to hold under the same interest rate, which stimulates the demand and pushes the interest rate down. The algorithm reacts in the way described above balancing the supply against the growing demand.
Liquidity Trap
The interest rate has a lower boundary (Imin) at which the difference between staking coins and keeping them in the free form becomes negligible for most economic agents. If the interest rate reaches the said boundary, the stake pool will become inelastic to the market conditions, and our policy will lose its efficiency. This situation resembles the Keynesian concept of a liquidity trap.
To counter a liquidity trap, we modify the rule and depeg the issuance rate from the interest rate. If I<Imin, the algorithm starts to increase ki exponentially until I=Imin. Since ks is inversely related to ki during the expansion, it will decrease simultaneously thus increasing the helicopter share. These measures will continue the expansionary policy until an equilibrium is found at S*. Eventually, with the help of helicopter coins, the upward pressure created by the described measures will push I back to the point at which the general policy rule can be applied again.
Contractionary Policy
The first stage of the contractionary policy is executed according to the policy rule in the way as described above but by decreasing ki instead of increasing it. ki, however, can only be non-negative. There is also a lower boundary (kimin) crossing which we can jeopardize the security of the blockchain. This value will correspond to a respective value of the interest rate (Ic) at which we stop decreasing ki and start increasing kc, which was set to 0 previously.
Under I>Ic, the algorithm adjusts kc according to the general rule kc=2(4-I)+5-kimin and shifts S* proportionally. Stage two contraction is performed not via the adjustment of the supply but via the adjustment of the velocity of circulation: by increasing ks and S* the algorithm makes the stake pool grow, which reduces the share of freely circulating coins (1-S) and creates a downward pressure on I. Eventually, the escalated pressure will push I back to Ic, when the algorithm drops kc back to 0 and returns to adjusting ki.
The bulk of this article was originally posted on Medium.
Long ago, when I found out about Bitcoin, I was very excited about the concept of a new generation of money powered by a decentralized ledger technology (just like everyone of us was, I suppose). The time has shown, however, that Bitcoin has failed to become a monetary asset and turned into a hedging tool or a speculative asset instead.
In 2018, after completely realizing that Bitcoin wouldn’t substitute fiat money, I began working on a cryptocurrency that would succeed in reaching that goal. I had to develop a completely novel consensus protocol from scratch, which was an extremely challenging task, but the most challenging issue was the development of an appropriate economic model.
Meanwhile, some developers of algorithmic stablecoins tried to solve the same issue with quite underwhelming results (Hello, UST!). The mere idea of stablecoins is still inappropriate for an asset that can become a new generation of money and substitute fiat because a derivative of an asset cannot become a full-scale alternative to the said asset, not speaking of depegging risks and the vulnerability to speculative attacks.
After all, to create the money of the future, we need to completely detach decentralized currencies from their fiat counterparts by upgrading the concept of a stablecoin to the concept of a stable economy, the difference being wider capabilities of the algorithm: instead of simply reacting to shifts in demand against a particular asset to maintain a preset exchange rate, it should be capable to counter shifts in demand coming from all possible channels and directions and maintain targeted values of the variables used in conventional macroeconomic policies, namely inflation and/or output growth.
Simply put, to create a decentralized money we need to create a decentralized Fed or, as I call it, a Decentralized Algorithmic Central Bank (DACB). Apparently, we cannot adapt the Fed’s discrete monetary policy since discretion is incompatible with decentralization, which is why DACB is inherently a model with an incorporated feedback policy rule.
Herein, I present the economic model of a Decentralized Algorithmic Central Bank, which allows targeting inflation via a feedback policy rule similar to the Taylor and McCallum rules. Unlike stablecoins, which are only capable of maintaining the exchange rate against a particular asset, the model performs as a fully-fledged central bank and is a generation ahead of all stablecoin concepts. Besides, the model doesn't fall under the stablecoin definition within MICA regulations and operates with native utility coins of the system, so the security regulations are also circumvented, which is a huge advantage over any existing stablecoin.
DACB VS Fed
Simply put, we can describe the Fed’s discrete policy as targeting inflation and/or employment via the control of the short-term interest rate, which serves as an intermediary target. The Fed uses four main policy tools to keep the interest rate within the targeted boundaries: the interest on reserves and ON RRP rate to set the lower boundary, the discount rate to set the upper boundary, and open market operations to adjust the available reserves in the banking system and keep them sufficient.
My DACB implementation directly targets the short-term nominal interest rate, which also allows to indirectly target other economic variables that are correlated with the interest rate, the primary one being inflation via the Fisher effect. To control the target, the algorithm uses the adjustment of the monetary base and velocity of circulation via the policy tools.
The issuance rate is the main policy tool. By increasing it, the algorithm performs the expansionary measures pushing the targeted value upward. By decreasing it, the algorithm executes stage one contractionary policy pushing the targeted value downward.
The contractionary rate is the secondary policy tool. It is activated during stage two contraction, when the issuance rate reaches its lower boundary. Stage two contraction reduces the velocity of circulation by increasing the stake pool, which pushes the targeted value downward.
Features of DACB Model
DACB Incorporates Feedback Policy Rule
DACB is a generation ahead of stablecoins, for it can not only follow the exogenous course by being pegged to allegedly stable off-chain assets but sustain a targeted equilibrium endogenously.
DACB incorporates a variation of a feedback policy rule with a targeted interest rate similar to the Taylor or McCallum rules to create a functional stable economy that satisfies the needs of all participants. With such a functionality, the model can make Dynemix the money of the future.
DACB Is Independent and Decentralized
All stablecoin designs feature a common downside: they always rely on external sources and are pegged to exogenous assets.
The DACB model makes a major breakthrough by introducing a concept that relies exclusively on on-chain data, is fully algorithmic, and is not related to any particular off-chain asset (USD, gold, SDR etc.) Instead, the model operates in conjunction with the entire global economy.
DACB Works with Native Coins
Algorithmic seigniorage-style stablecoins are paired with special assets to balance the supply against the demand.
The design of such assets endows them with the properties of securities, which puts the system under the control of regulators, creates significant legal obstructions, and severely undermines decentralization.
DACB operates directly with native utility coins, which allows to avoid all the respective difficulties occurring with the introduction of ancillary assets, including sanctions from financial regulators. For example, it’s not classified as a stablecoin (neither ART nor EMT) under the MICA regulation or as a security token under MiFID II.
DACB Is Highly Egalitarian
With the help of helicopter coins, the model impedes the concentration of power and doesn’t allow strong actors to sustain their influence endlessly simply by staking all their coins and increasing the relative share in possession.
Since banks and credit are removed from the model, strong financial actors can no longer take direct advantage of a monetary expansion and receive or create money out of thin air.
Helicopter coins also stimulate the demand from the consumer side, which allows to circumvent many credit-related monetary issues.
DACB Can Be Potentially More Efficient Than Fed
The effectiveness of a monetary policy is fundamentally limited by the credibility of a monetary authority. In practice, credibility can be a weak point of a monetary policy. Despite being commonly positioned as independent actors, CBs often suffer from governmental interference, which can undermine credibility and render applied measures less effective. The governing boards of CBs can also be affiliated with commercial banks, which further aggravates the issue.
The DACB model is deprived of that flaw for it is completely decentralized and trustless. There can be no question of credibility or impartiality of an algorithm that is incorporated into a blockchain system, which makes the monetary policy highly efficient in terms of managing expectations and driving the behavior of participants in a desired direction.
DACB Can Seamlessly Operate with Conventional and Islamic Banking
Most of the countries that follow Islamic financial doctrine suffer from severe economic difficulties for Islamic banking is not consistent with conventional banking, which is based on credit.
Since my model doesn’t rely on credit, it is fully compliant with the Islamic principles and can be adopted on the full scale by Muslim countries, thus solving the problem of seamless integration of the Islamic banking into the global financial infrastructure and allowing to overcome economic difficulties that many of Muslim countries are currently facing.
Technical description
Here's some boring technical crap that nobody will probably read, but I'll still leave it here just in case. If you are into it, though, you can go straight to the full paper.
DACB, Dynemix Blockchain and Liberdyne Messenger
Dynemix is a decentralized, permissionless, account-based blockchain system powered by a unique Proof-of-Stake BFT consensus protocol. Dynemix is the first DLT system to incorporate a DACB model. The DACB model of my design requires a number of specific technical features from the underlying blockchain, which is why it cannot be implemented into any existing platform, and which is why I do not give it a separate name and simply call it “the economic model of Dynemix”.
Dynemix features a low-level transport protocol for the delivery of messages to offline users called DOG. The Liberdyne messenger incorporates the said protocol in its design, and thus provides a practical solution for the distribution of helicopter coins, which are a crucial element of my DACB’s design.
Helicopter Coins
Current gen blockchains (and other DLT designs) feature a common coin distribution model: newly mined (minted) coins are distributed among block producers as an incentive to process state transition and keep the system secure.
Dynemix introduces a novel approach: only a share of newly issued coins goes to the minters of the block, whereas the rest are dispersed among users of the platform, thus becoming an embodiment of the concept of helicopter money introduced by Milton Friedman.
The share of coins dispatched to minters called a minters’ share can be arbitrarily adjusted by the algorithm within the predefined boundaries, which serves as one of the tools of the monetary policy, as it, together with the issuance rate, defines the nominal interest for staking.
Tools of Monetary Policy
The algorithm conducts the monetary policy via the following tools:
Issuance rate (ki) – the annual rate at which the coin supply grows.
Minters’ share (ks) – the share of newly issued coins that is dispatched to minters as a reward for a block creation.
Contractionary rate (kc) – the annual rate at which the targeted size of the stake pool (S*) is being increased.
Policy Rule
The algorithm conducts the monetary policy according to the following general rule:
ki+kc = 2(I*-I)+ki*
Where:
ki is the issuance rate;
kc is the contractionary rate;
I is the current nominal interest rate;
I* is the targeted nominal interest rate;
ki* is an assumed issuance rate that maintains the targeted inflation rate and potential GDP.
Suppose we agreed that the potential GDP growth rate is 3%, the targeted inflation rate is 2% and the natural real interest rate is 2%. Then we can rewrite the equation as follows:
ki+kc = 2(4-I)+5
Default Equilibrium
If we target economic variables to reach the values stated above, we should set the algorithm as follows: the issuance rate should be set to 5% to provide 2% inflation assuming 3% economic growth; to target 4% nominal interest (which equals to 2% real interest under 2% inflation) and 20% stake pool size (which provides enough security), the minters’ share should be set to 16%.
Under the stated setup, the interest rate curve would look as follows:
We can see that with the growth of the stake pool (S), the interest rate for stake holders (minters) declines. As more coins are locked as stakes, it becomes less profitable to participate in minting, and some of stakeholders will unstake their coins to invest them into higher yielding assets. There is a point of equilibrium at which S holds at any given time. This point represents the current interest rate in the economy. On the chart above the interest rate is 2% in real terms or 4% in nominal terms, which is the targeted r* for the algorithm.
A change in the macroeconomic setup will cause a respective change in the desired interest rate, and the size of the stake pool will shift either upward, which will trigger the expansionary policy, or downward, which will trigger the contractionary policy.
Expansionary Policy
Suppose the inflation rate is expected to drop from 2% to 1%. According to the Fisher effect, I will drop to 3%, which will push the equilibrium to S=26.6. Under I<I*, kc=0, and the algorithm will recalculate ki according to the rule above: ki=2(4-3)+5=7. ks will be also adjusted to offset changes in S*. Therefore, for 1% expected drop in inflation the algorithm increases the issuance rate by 2%, which provides a counterpressure on the interest rate and pushes the system back to the optimal equilibrium. The coefficient of the rule (a) can be arbitrarily adjusted. I suggest a=2 for the general rule as a more aggressive reaction than proposed by Taylor (1993), namely a=1.5, due to the specifics of the blockchain setting.
The same reaction will be applied to the pressure on the interest rate coming from other channels. For example, if the exchange rate of our coins against exogenous assets (say USD) is expected to shift upwards, our currency becomes more attractive to hold under the same interest rate, which stimulates the demand and pushes the interest rate down. The algorithm reacts in the way described above balancing the supply against the growing demand.
Liquidity Trap
The interest rate has a lower boundary (Imin) at which the difference between staking coins and keeping them in the free form becomes negligible for most economic agents. If the interest rate reaches the said boundary, the stake pool will become inelastic to the market conditions, and our policy will lose its efficiency. This situation resembles the Keynesian concept of a liquidity trap.
To counter a liquidity trap, we modify the rule and depeg the issuance rate from the interest rate. If I<Imin, the algorithm starts to increase ki exponentially until I=Imin. Since ks is inversely related to ki during the expansion, it will decrease simultaneously thus increasing the helicopter share. These measures will continue the expansionary policy until an equilibrium is found at S*. Eventually, with the help of helicopter coins, the upward pressure created by the described measures will push I back to the point at which the general policy rule can be applied again.
Contractionary Policy
The first stage of the contractionary policy is executed according to the policy rule in the way as described above but by decreasing ki instead of increasing it. ki, however, can only be non-negative. There is also a lower boundary (kimin) crossing which we can jeopardize the security of the blockchain. This value will correspond to a respective value of the interest rate (Ic) at which we stop decreasing ki and start increasing kc, which was set to 0 previously.
Under I>Ic, the algorithm adjusts kc according to the general rule kc=2(4-I)+5-kimin and shifts S* proportionally. Stage two contraction is performed not via the adjustment of the supply but via the adjustment of the velocity of circulation: by increasing ks and S* the algorithm makes the stake pool grow, which reduces the share of freely circulating coins (1-S) and creates a downward pressure on I. Eventually, the escalated pressure will push I back to Ic, when the algorithm drops kc back to 0 and returns to adjusting ki.
The bulk of this article was originally posted on Medium.