# Running amm_arb strategy¶

Currently the gateway V2 only supports the amm_arb strategy. A good way to safely test amm_arb strategy is to set it up on Kovan testnet, and arbitrage against Binance in paper trading mode.

Running integration tests with this configuration has the following advantages:

1. Kovan test tokens are easy to obtain - just go to https://faucet.paradigm.xyz/.
2. No actual CEX account, or CEX balance is required to perform testing.

## Obtaining Testnet Tokens¶

If you’re testing on Kovan, go to https://faucet.paradigm.xyz/. You’ll be provided with the following testnet tokens:

• 1 ETH
• 1 WETH
• 500 DAI

If you’re testing on Avalanche, go to https://faucet.avax-test.network/. You’ll be provided with 2 testnet AVAX.

### Connecting Wallets¶

Start Hummingbot, and issue the following commands to connect to your test wallet.

# Connecting Ethereum wallet to Uniswap protocol
>>> gateway connect uniswap

# Connecting AVAX wallet to Pangolin protocol
>>> gateway connect pangolin


Remove private keys from clipboard

Make sure to remove the copied private keys from your clipboard after use.

## Setting Up Global Configuration¶

You’ll need to enable debug console to allow triggering arbitrage transactions later. Also, you should set up some appropriate paper trade account balance to allow amm_arb to trade on both sides.

e.g. you’re going to set up amm_arb to trade between WETH-DAI Uniswap / Kovan vs. ETH-USDT on Binance paper trade. Then, you’ll need to set up some initial ETH and USDT balance on the paper trade account in conf_client.yml . Here’s an example setup you can use for the global config.

debug_console: true
ETH: 10
USDT: 30000


## Setting Up amm_arb Strategy¶

Use create command to set up an amm_arb strategy. Make sure you’re trading between a test-net gateway connector, vs. a CEX in paper trading mode. Also, use small order amounts so you don’t burn up all your test tokens in just one order.

Here’s an example config:

Connector 1: uniswap_ethereum_kovan
Market 1: WETH-DAI
Market 2: ETH-USDT
Order amount: 0.1
Min profitability: 1
Market 1 slippage buffer: 0.05
Market 2 slippage buffer: 0
Concurrent orders submission: choose any


After creating the config, use the config command to set debug_price_shim to True.

>>> config debug_price_shim True


## Starting amm_arb Strategy¶

Use the start command, and then answer Yes when asked to confirm the strategy settings.

After starting the amm_arb strategy, you’ll see the apparent prices on the AMM market will track the prices on the CEX exchange - and if there is none, the strategy will report no arbitrage opportunities.

## Testing on Pangolin¶

Testing on Pangolin needs to be done on Avalanche mainnet. There is no easy way to test Pangolin on Avalanche’s Fuji testnet. This implies real Avalanche tokens will be needed to perform tests.

Below are some recommended steps for performing tests on Pangolin.

Preparations

1. Buy 2 AVAX from a real exchange, and send to test wallet.
2. Go to the Pangolin Exchange website, and wrap 0.8 AVAX into WAVAX.
3. Trade 0.8 AVAX into USDT.e
4. In Hummingbot, issue gateway connect pangolin, and connect using the private key of the test wallet. Remember to clear out your clipboard by copying some other stuff into it, after pasting the private key.
5. Add some entries AVAX and USDT to the paper trading balance in conf_client.yml.

Strategy Setup

1. Set up an amm_arb strategy to trade between Pangolin on Avalanche, vs. Binance paper trade. Use WAVAX-USDT.e as the trading pair on Pangolin, and AVAX-USDT as the trading pair on Binance paper trade. Set a reasonably small order size to avoid burning through your wallet balance in one trade. e.g.

Connector 1: pangolin_avalanche_avalanche
Market 1: WAVAX-USDT.e

2. Enable debug price shim in the strategy config by issuing config debug_price_shim True in Hummingbot console.
4. Use GatewayPriceShim, as detailed in the last section, to trigger arbitrage opportunities on the AMM side. Feel free to use relatively large price deltas (e.g. Decimal(40)) to trigger arbitrage orders - because small order sizes would require a large price delta to cover gas costs.