Comment on page


Nakō: The first DeFi native hedge for FX

The current nightmare to hedge FX exposure

Hedging non-native assets in DeFi has always been a challenge. If you try to do it on-chain there is always a lack of liquidity, which equates to higher costs. If you do it off-chain, there are inefficiencies because your assets are on-chain plus the entire burden of managing risk with a bank as a counterpart.
While the former is harder to equate and is actually not always possible (how many "safe" venues allow you to trade emerging market currencies on-chain?) the latter has been used extensively and the costs are obvious. Lets take the Brazilian Real as an example of the latter.
  • The investor holds 100 USDC and wants to invest in BRZ (BRL stable coin) denominated assets, yielding 15% p.a..
  • He would first trade his USDC for BRZ, lets say at at relatively liquid venue he'll receive a 0.50% mark-up to official spot. He is left with 99.50 of the equivalent USDC value in BRZ.
  • He then deploys into the protocol holding the asset and yield.
  • In parallel he sells BRL futures (NDFs) with his local bank, to hedge exposure to BRL's volatility. The NDF costs him the rate differential between BRL treasuries and USD treasuries, currently at around 10%. Since banks don't work for free they'll charge him an extra 0.50% commission. For a total hedge cost of 10.5% p.a.
    • Our friend's rate has already moved down from 15% to 4%, and it gets worst.
  • The bank will demand collateral to guaranty our investor's ability to honor his side of the NDF, for emerging market currencies the collateral is usually 10%-20% of the notional amount. Plus it needs to be posted in a real USD account, forget about USDC. So here our investor will have to reduce his effective allocation to the strategy in order to post collateral.
    • The bank will hold 10% of the $99.50 so he is left with $89.55 to invest. Taking his effective rate to 3.5%
  • And finally, he receives his invested BRZ back, swaps it back to USDC, taking another 0.50% hit and is left with a final effective rate of 3%
As you can see our friend starts off like this 🤑 and ends up like this 🤕, and this was the best case scenario. If he is a small client for sure commissions would be higher, and if there is a strong movement in the FX rate he would have to post more collateral.

Kona introduces Nakō, a new DeFi primitive for on-chain hedging

In the end of the day, a bank that provides an NDF is matching a client that wants exposure to a certain currency X with another client that wants exposure to currency Y. They don't need to match the exact flow as these banks usually have cross border books to match excess demand and supply. We use this deconstructed model of an NDF to build Nakō.
The central piece to understanding how our solution works is to accept that there will always be users that want risk in different currencies. These users either want to speculate (have exposure to the currency's movement) or have a fundamental exposure (imagine you are living in Brazil and all your monthly costs are in BRL).
The second important concept is simple but many skip through it without noticing. When you borrow a coin, you owe the quantity that you borrowed, not the financial amount (quantity x price). So if you borrow 10 ETH you will owe 10 ETH, it doesn't matter if ETH is priced at $1,000 or $5,000. What will change is your required collateral. So if you borrow you're neither short nor long.
By combining both points we arrive to a simple but elegant solution, the fusion of an AMM with a Lending pool. A single pool where liquidity providers with a fundamental need for currency X, can lend to investors that want access to strategies that are exposed to currency X by providing collateral in their native currency Y.
LPs carry the volatility and are paid for their loans, the pool holds collateral until the loan is paid, and investors are exposed to the clean yield.
This mechanism can be applied as a closed loop within the protocol, where 100% LTV would suffice, as well as offered as a service to other protocols. Furthermore, it could be used as a standalone function where different risk parameters would have to be developed.

How would this work

  1. 1.
    A USDC pool is deployed by Kona, and is funded by BRZ LPs
  2. 2.
    Investors allocate the desired amount of USDC to the strategy
  3. 3.
    The pool locks USDC as collateral and transfers BRZ from its treasury to the strategy
  4. 4.
    The strategy matures and returns the BRZ to the pool
  5. 5.
    The pool unlocks the USDC held plus interests from the strategy minus interest to LPs
    • If there is a loss in the strategy, the pool holds on to the collateral equivalent to the loss amount and transfers the net funds back to investors.