Thread by @rektdiomedes

Thread: Should You Worry About Impermanent Loss?

Fun pop quiz, which of the following would have been profitable (vs just holding) if taken 6 mo ago and held til today?

a) ETH-AAVE at 30% apr
b) ETH-FTM at 50%
c) ETH-USDC at 20%
d) AVAX-USDC at 100%
e) AVAX-USDC at 250%

1/x

Answers to quiz at bottom of thread :) But first, a brief breakdown of Impermanent Loss (hereafter called 'IL').

Since I started farming fulltime, IL has occupied my thoughts every day, but its a tricky subject, and there's some bad takes you'll see floating around on it.

2/x

First a definition: IL is the relative loss you incur from providing tokens in a liquidity pool within an AMM, vs just holding them for (hopeful) appreciation.

There is always IL, the only question is if the yield you are given as an incentive to LP will outweigh the IL.

3/x

There is no such thing as 'Impermanent Gain', a phrase you sometimes seen thrown around.

The best case scenario LP'ing (not counting yield) is that the ratio of the two tokens you're LP'ing stays the same, and you have 0% IL.

Any divergence and you have IL.

4/x

That's how liquidity pools on AMM's work.

A useful heuristic I've come up with for my own use is this:

The value of an LP position will always average towards the worst performing of the two tokens.

5/x

It doesn't matter if Token A and Token B both go up, both go down, A goes up and B goes down, or vice versa.

If the ratio between them changes at all, your LP value will average towards the poorer performer.

6/x

Obviously this is a major thing to consider when evaluating yield farming opportunities.

Luckily, there are some great calculators we can use to gauge the risk of IL relative to farming incentives.

7/x

Dailydefi (dot) org/tools/impermanent-loss-calculator/ is probably the most commonly one used.

It does not factor in yield (apr) though, so if you want that included, check out the amazing work of @phtevenstrong, who created an IL calculator that DOES also include yield.

8/x

As the calculators show us, if we LP two tokens that are both worth $1.00, and Token A eventually gets to $2.00 and Token B stays at $1.00 (a divergence to 2x), IL would be 5.72%.

If the tokens go from $1/$1 to $5.00/$0.50 (a divergence to 10x) the IL would be 42.5%.

9/x

The question is thus whether the yield we get as an incentive for LP'ing will outweigh the IL and let us generate a meaningful profit.

Let's get back to the 'Pop Quiz' examples from the start :)

10/x

Again, which of the following would have been profitable (vs just holding) if taken 6 mo ago and held til today?

a) ETH-AAVE at 30% apr
b) ETH-FTM at 50%
c) ETH-USDC at 20%
d) AVAX-USDC at 100%
e) AVAX-USDC at 250%

11/x

If you guessed a, c, and e were the profitable ones where farming was worth it, you're right :)

Let's break them down one by one...

12/x

a) ETH/AAVE 30% APR pair went from $2279.35/$250.66 to $4075.03/$279.04.

$1000 split between them would be worth $1450.51
today if not LP'd, $1410.75 if LP'd (2.74% IL), but $1591.56 w/the yield.

= 🐸😀

13/x

b) ETH/FTM 50% APR pair went from $2279.35/$0.24 to $4075.03/$2.39.

$1000 split between them would be worth $5873.07
today if not LP'd, $4219.42 if LP'd (28.16% IL), and $4871.85 w/the yield.

= 🐻🙁

14/x

c) ETH/USDC 20% APR pair went from $2279.35/$1.00 to $4075.03/$1.00.

$1000 split between them would be worth $1393.09
today if not LP'd, $1337.09 if LP'd (4.08% IL), and $1453.94 w/the yield.

= 🐸😀

15/x

d) AVAX/USDC 100% apr pair went from $11.98/$1.00 to $115.33/$1.00.

$1000 split between them would be worth $5313.44 if not LP'd, $3102.72 if LP'd (41.61% IL), and $4128.40 w/the yield.

=🐻🙁

16/x

e) AVAX/USDC 250% apr pair also went from $11.98/$100 to $115.33/$1.00.

$1000 split between them would (again) be worth $5313.44 if not LP'd, $3102.72 if LP'd (41.61% IL), but $5666.92 w/the yield!

=🐸😀

16/x

ETH-AAVE and ETH-USDC were profitable despite the low APR's, bc the ratio between their prices stayed so static.

ETH-FTM's modest APR couldn't overcome the IL that arose from FTM mooning.

The same applied w/AVAX-USDC at 100%, though 250% was enough to overcome the IL :)

17/x

This goes to show that if you are super bullish on a token, you may want stick to single-sided opportunities instead of LP'ing w/ it.

You never know though what will happen in the future, and therefore I think LP'ing is still a fantastic strategy (if w/sufficient APR's!).

18/

What I like to do is mix up the strategies, and do some LP'ing as well as some single-sided farming/staking/lending/etc.

This way I'll end up generating substantial returns no matter what the token ends up doing.

19/x

You can also try to pair tokens that behave similarly. For instance ETH/BTC or ETH/AAVE.

You definitely want to remember that the apr you're getting is a YEAR of apr though, so often its better to look at the yield you'll earn daily or monthly instead.

20/x

Also, a note on math..

There's many different ways to calculate all this.

My calculations did NOT include autocompounding, like you would get if you use the wonderful @beefyfinance :)

Also, (cont)...

21/x

I applied the 6 mo's of APR (50% of stated apr) to the midway point of the LP's change in value, vs applying it to the end value (since the token prices go up gradually).

For this I just used the dailydefi calc + the above apr multiplier.

22/x

Finally, if you're interested in yield farming + similar, follow these awesome accounts!

@TaikiMaeda2
@rektfoodfarmer
@noahseidman
@Route2FI
@takegreenpill
@CFrogE1
@blocmatesdotcom
@chng_raymond
@Rewkang
@Ceazor7
@korpi87
@benblute
@Cryptoyieldinfo

23/x

If you found this interesting please retweet and follow! :)