Skip to content

Commit

Permalink
Merge pull request #20 from geniusyield/19-mm-review-readme
Browse files Browse the repository at this point in the history
19 mm review readme
  • Loading branch information
sourabhxyz authored Jan 4, 2024
2 parents 12720fc + e090b42 commit 7757d66
Showing 1 changed file with 28 additions and 18 deletions.
46 changes: 28 additions & 18 deletions README.md
Original file line number Diff line number Diff line change
@@ -1,17 +1,17 @@
# Market maker bot

> **⚠️ Warning**
>
> [!WARNING]
> Market making is a risky activity and running this bot can lead to loss of funds.
Market maker bot for [GeniusYield](https://www.geniusyield.co/) DEX which implements _fixed spread versus market price strategy_.

## Fixed spread vs market price strategy

> **ⓘ Order classification and price**
> [!NOTE]
> **Order classification and price**
>
> We call non-ada tokens as _commodity_ and ada as _currency_. Order offering currency in exchange of commodity is called as _buy order_ whereas order offering commodity in exchange of currency is called as _sell order_.
>
> We call non-ada token as _commodity_ and ada as _currency_. Order offering currency in exchange of commodity is called as _buy order_ whereas order offering commodity in exchange of currency is called as _sell order_.
>
> _Price_ is described in display unit[^1] of currency token per display unit of commodity token.
Given a market price `M` and a variable `δ` defined as _spread_, bot would place following orders where exact number and volume is determined by configuration:
Expand All @@ -27,13 +27,13 @@ Given a market price `M` and a variable `δ` defined as _spread_, bot would plac
* `M * (1 + δ + δ / 2 + δ / 2)`
* And so on, where `n`th sell order's price is given by `M * (1 + δ + (n - 1) * δ / 2)`.

If market price has drifted way higher (_"way higher"_ as directed by configuration) than the price at which buy orders were placed, buy orders would be canceled. Likewise, if price has drifted way lower than the price at which sell orders were placed, they would be canceled.
If market price has drifted way higher (_"way higher"_ as directed by configuration) than the price at which buy orders were placed, buy orders would be canceled. Likewise, if the price has drifted way lower than the price at which sell orders were placed, they would be canceled.

## Running the market maker bot: Using docker compose
## Running the market maker bot: Using docker compose (simple)

The simplest way to start an MM bot instance is by using Docker compose.

After cloning the repository a few environment variables must be set. After this has been done; the MM bot container can be started using `docker compose`:
After cloning the repository only a few environment variables must be set. As soon as this has been done; the Market Maker can be started using `docker compose`:

``` bash
# Clone the repository:
Expand All @@ -50,31 +50,37 @@ docker compose up
As in the example above; the following environment variables must be specified before calling `docker compose up`:
- `MAESTRO_API_KEY`: The MAINNET API key to be used for accessing the Maestro services.
- `PAYMENT_SIGNING_KEY`: The payment signing key to be used. Please see the [signing key generator](https://github.com/geniusyield/signing-key-generator) for details.
- `COLLATERAL_UTXO`: A suitable UTxO with 5 ADA to be used as colletaral UTxO.
- `COLLATERAL_UTXO`: A suitable UTxO with 5 ADA to be used as collateral UTxO.

The configuration values used for these environment variables in the example above are just placeholders. These must be replaced by your own
configuration values. A MAINNET Maestro API key is needed, a payment signing key must be generated and a collateral UTxO must be provided after
sending funds to the address controlled by the payment signing key.

Maestro API keys are available after registration via the following link:
- https://docs.gomaestro.org/Getting-started/Sign-up-login

> [!WARNING]
> Please make sure to adapt the `MARKET_MAKER_CONFIG` configuration according to your needs! Please see the docker-compose.yml file for further details.
> Please make sure to adapt the `MARKET_MAKER_CONFIG` configuration according to your needs! Please see the docker-compose.yml file for further details. For the configuration of the Market Maker, please see the [Configuration Settings](#Configuration) chapter.
## Running the market maker bot: Building from source (advanced)

In case you would like to build the Market Maker Bot from source, this chapter covers how to accomplish this.

## Running the market maker bot: Building from source
> [!TIP]
> If you are not planning to contribute to the project, simply using the pre-built docker image, as described above, is likely the easier way to get started.
First, you need to setup the necessary tooling to work with [haskell.nix](https://github.com/input-output-hk/haskell.nix), then simply run `nix develop`, and it will drop you into a shell with all the necessary tools. Once inside the environment, you can build the order bot with `cabal build all`.

> ****
>
> [!NOTE]
> Nix is not necessary if your environment already has the right set of dependencies. One may look at the [CI file](https://github.com/geniusyield/atlas/blob/main/.github/workflows/haskell.yml) for our transaction building tool, which current project also relies on, to see dependencies used.
Then the bot can be ran with following command: `cabal run geniusyield-market-maker-exe -- Run my-atlas-config.json my-maker-bot-config.json` where `my-atlas-config.json` is the configuration for [Atlas](https://github.com/geniusyield/atlas) and `my-maker-bot-config.json` is the configuration of our market maker bot.

See [`atlas-config-maestro.json`](./atlas-config-maestro.json) & [`atlas-config-kupo.json`](./atlas-config-kupo.json) as an example of Atlas configuration using [Maestro](https://www.gomaestro.org/) provider & local node with [Kupo](https://github.com/CardanoSolutions/kupo) respectively.

### Bot configuration
### Configuration

> ****
>
> [!NOTE]
> See [`sample-preprod-maker-bot-config-gens.json`](./sample-preprod-maker-bot-config-gens.json) and [`sample-mainnet-maker-bot-config-gens.json`](./sample-mainnet-maker-bot-config-gens.json) for sample Preprod and Mainnet market maker bot configuration respectively.
```json
Expand Down Expand Up @@ -145,10 +151,14 @@ See [`atlas-config-maestro.json`](./atlas-config-maestro.json) & [`atlas-config-

## Canceling all the orders

If you want to cancel orders placed by the simulator you can run `cabal run geniusyield-market-maker-exe -- Cancel my-atlas-config.json my-maker-bot-config.json`.
If you would like to cancel *ALL* orders placed by your Market Maker Instance, you can simply run:

``` bash
cabal run geniusyield-market-maker-exe -- Cancel my-atlas-config.json my-maker-bot-config.json
```

## Known Issues

* When bot tries to place multiple orders in a single iteration, it might happen that we pick same UTxO against different transaction skeletons (due to a [quirk](https://github.com/geniusyield/dex-contracts-api/blob/cf360d6c1db8185b646a34ed8f6bb330c23774bb/src/GeniusYield/Api/Dex/PartialOrder.hs#L489-L498) where place order operation specifies UTxO to be spent in skeleton itself), leading to successful building of only some of the transaction skeletons and thus only few of the orders might be successfully placed even though bot might very well have the required funds to place all. Now bot can place remaining ones in next iteration but as of now, these next orders are placed starting with initial spread difference from market price leading to a situation where bot might have multiple orders at the same price.
* When bot tries to place multiple orders in a single iteration, it might happen that we pick same UTxO against different transaction skeletons (due to a [quirk](https://github.com/geniusyield/dex-contracts-api/blob/cf360d6c1db8185b646a34ed8f6bb330c23774bb/src/GeniusYield/Api/Dex/PartialOrder.hs#L489-L498) where place order operation specifies UTxO to be spent in skeleton itself), leading to successful building of only some of the transaction skeletons and thus only few of the orders might be successfully placed even though bot might very well have the required funds to place all. Now the bot can place the remaining ones in next the iteration but as of now, these next orders are placed starting with initial spread difference from market price leading to a situation where the bot might have multiple orders at the same price.

[^1]: _Display unit_ is one to which decimals are added as directed under [`cardano-token-registry`](https://github.com/cardano-foundation/cardano-token-registry).

0 comments on commit 7757d66

Please sign in to comment.