Is impermanent loss really a loss?

I’m trying to wrap my head around this and realizing I won’t get the number of tokens back compared to the amount I provided for liquidity. But if I’m staking a token with USDC and token goes up in price is it really a loss?? I get less of the token back but it’s still worth more. So it seems like I’m still making money. Am I just making less money than I could have made? Can I actually lose money in terms of fiat? or is it just like losing value in terms of that number of token? If I stake a token 1 to $1 USDC, and token goes to $1000, how am I really losing?

View Source

36 thoughts on “Is impermanent loss really a loss?”

  1. A more accurate name would be “Reduced gain”. It’s really not as bad as it sounds if you do the math. If a coin doubles in value, the IL is only 5% assuming the other coins value is stable. This is why I only LP stable/bluechip coins, as those bluechip coins are unlikely to moon/crash and cause massive IL. By virtue of being a bluechip, everybody wants them. High volume = more profits.

    I dislike variable/variable pairs due to the uncertainty. Sure, some coins tend to move together… but what if they don’t? I prefer the reduced risk (and potentially reduced profit) of stable/variable pairs.

    IL asserts that in the LP vs HODL scenario the hodler profits more when selling the assets than the LPer does when withdrawing from the pool and selling. The thing is, I’m not selling my LP coins anytime soon. Crypto is volatile and will continue going up and down. I’ll happily collect fees that entire time and tolerate some IL, while a hodler earns **nothing**. To me, it’s a no brainer to LP a stable/bluechip.

  2. Hi,

    Impermanent loss is the loss versus HODL a 50/50 portfolio. So if you have two tokens ETH and USDC for exemple, if ETH goes up you would make money but not as much as if you had held 50% ETH and 50% USDC.

    If ETH loses money, you will lose money in USDC terms and more than if you had held a 50/50 portfolio.

  3. IMO the naming is terrible. It should just be called impermanence and drop the ‘loss’ part which people get hung up on.

    The loss is really the ‘what if’ part, as in what if I had done this, but thats a terrible way to live life. I’ve heard IL described as the cost of doing business which I like better.

  4. I think you have the concept of impermanent loss backwards and most importantly with LP pairs you have to consider the summation of the total you provided, not just the gross/increase in tokens for a particular half. Yes you would be losing money in terms of fiat and most importantly if a token does go up in price depending on the APR of the pool, you would most likely receive more of the half that is worth less. Example: ftm-usdc, u receive more usdc if ftm does go up respective of percentage increase. Honestly using apeboard which tracks a number of protocols allows for easier understanding.

  5. As for me, I think single asset staking is better than LP ,because the only loss you will incur is when the token dips but some projects like blockbank and Celsius allows stable coins staking.

  6. Yes, in the way opportunity cost is a real cost. You’re not spending out of pocket but it sucks to know you shouldn’t have bothered with LPs.

    No, in the eyes of the tax man. It’s a loss only when you sell for less than you paid for it.

  7. You lose money vs. holding the original ratio of the two tokens. That’s before you factor in the LP rewards.

    It can potentially be a huge loss. It can also be nothing in comparison to the gains from providing liquidity.

  8. Best way to understand it is to just build a simple spreadsheet that calculates LP value based on the changing price of the assets. It’s very easy to see the effect it has. It’s basically just the effect on constant rebalancing, so you’re constantly selling your winning asset to buy your losing asset. It’s just a matter of doing the math to see where your breakeven is on asset price based on the expected APY of the pool.

  9. Impermanent loss is just a weird term to define the underperformance of your LP (that is always constantly rebalancing 50/50) compared to the same investment without rebalancing.

    Imagine you buy 50$ Google stock and 50$ FB. You do nothing and after 1 year Google triples and FB is flat. Now you have 150$ of Google and 50$ in FB.
    Your friend instead decided to keep rebalancing to 50/50 ratio so he constantly sold the winner stock and bot the flat stock, and the performance of both stocks didn’t swing up and down enough to earn some extra money trading in and out of each stock, so let’s say his portfolio is now worth 120$ (60$ in each stock). So he underperformed your portfolio, the “benchmark”.

    For some stupid reason people call this “impermanent loss” in the crypto world…

  10. I think of liquidity providing as a swap. You provide the pool with 2 assets once you take the liquidity out of the pool you end up with more or less of one of the assets. You just need to figure out what you want to accomplish with lp. For example, you provide liquidity to the usdt/usdc pair 50/50 then, once you take the liquidity out you ended up with 30 usdt and 70 usdc. So, basically you exchanged 20 usdt for 20 usdc.

  11. >But if I’m staking a token with USDC and token goes up in price is it really a loss??

    Nope. You’re still up–not as much as you would be if you just held the token, but you didn’t take as much risk to get there.

    >Am I just making less money than I could have made?


    >Can I actually lose money in terms of fiat?

    In stable/unstable pairs, if the unstable goes up and the stable stays… stable, you can’t lose fiat. You merely don’t make as much as if you held the coins separately…

    …except that you’re also making fees, which if you picked a decent exchange, will usually outweigh your IL.

    In your example, where the unstable coin goes from $1 to $1000, impermanent loss is a big deal–you miss out on 93% of those gains. But the reality is, you *rarely* will see that kind of gain in a time period such that your fees won’t more than outweigh your impermanent loss.

  12. it is definitely a loss. If one pumps more than the other you get “less profit” but if you are in an LP where you have 100xx and 100MIM, and the xx token rugs and go to zero, you will actually end up with only 50MIM after withdrawing. This is worst case scenario but it actually happened to me and it is to make you understand it can be quite big

  13. I honestly don’t think it is a loss, unless you take that loss…..
    However, there are DEX’s that prevent or almost eliminate IL, like the AllianceBlockDEX

  14. I see a lot of people saying IL is like the cost of doing business or result versus holding the coins without the LP.

    The way I see it is, IL only affects you if you already HAD the coins and decided to add them to a liquidity pool. Then you would gain some % out of the LP fees (compared to you holding) and you would loose some % in IL (compared to you holding). In that case you would want for the fees earn to be greater than the IL.

    If you don’t have the coins beforehand, and decide to buy them JUST to add to a LP, the IL is a bit irrelevant as the holding scenario didn’t exist for you. In that case is only relevant how much you made out the whole “operation”.

  15. You could forget about IL if you just provide single asset to the LP, some protocols like Balancer allow that. This way you are not exposed to IL and you can double down on your token A

  16. I think the question should be, “Is impermanent loss really impermanent?”

    I have no idea how this concept was born. Yes, of course, if the prices of tokens in an LP swing back in “balance” you haven’t lost anything.

    But that’s the same of just holding a single token. If the price goes down and then goes back up, you no longer have a loss.

    With LPs you have other scenarios in which you can incur losses, but they are just as “temporary” as holding the tokens separately.

    A loss is a loss as soon as you close your position at a loss.

    “Impermanent loss” is a meaningless made-up term.

    I guess they wanted to label it a different way as well as marketing LPs in such a way that people will stay in them longer. That clearly hasn’t worked as people dry up all the incentives and move to the next higher APY LP.

    Funny enough though, because there are more ways to make a loss in an LP, the loss is more “permanent”, or more likely.

    For example if both your tokens go up in an LP but at a different rate, you still make a profit, just less than holding them separately. But in a way you made a “loss on your profit” (i.e. impermanent loss) a scenario that would not have presented itself if you just held the tokens.

    “LP loss” or “Balance Difference” would have been better and easier to understand.

    Obviously, I am just focusing on the “impermanent loss” part of LPs given the question of the OP, and I am not taking into consideration overall profits with fee earnings and incentives.

  17. That’s a real loss and even synthetic than having asset weaken in value. That’s why I take my time trading on Cex like Btay which is endeavouring to make thing transparent in this industry with its all in one platform featuring pro and newb traders, even nft collectors. You should switch method sometimes when liquidity mining doesn’t fit you well. Xd,

  18. No not really. It’s more “loss of potential gain over period” so your gain could’ve been higher if you just held the coin. But that’s not real loss.

  19. You can always lose money in terms of fiat, bro. But I believe that this “loss” exists only on paper really, since it will only happen if the LP withdraw their liquidity at a certain price. Remember also that when a fee is paid on a LP, the amount is distributed based on your share (it looks like a company paying dividends to its shareholders). This means that even though you get an impermanent loss (IL), the earnings you’ll get from providing liquidity tends to surpass and be more profitable than those profits of only hodling the assets.

    My experience with USDC is limited, since all of my position is deposited in Yield App. I can choose to receive all rewards in Yield (18%). But, bro.. IL will occur if the price moves in either direction. If you want to diminish IL’s risk, you can choose to receive part of the rewards in USDC itself (you’ll get a lower APY of course). What are you using for USDC? each protocol is different.

    Yield faming can also be a strategy to diminish the IL risk, since users can provide liquidity and earn money from fees and interest. DYOR for Venus, Kalmar and PancakeSwap. I’m testing the KalmySwap right now before its official launch and staking CAKE on Pancake. On Venus I’m just supplying assets by now. Anyway… IMHO anything is better than just hodling the assets.

  20. I was also confused about this. Have you clarified it?

    Because to me it seems that it’s not **really** a loss… In which case, I see no reason not to deploy **a lot** of capital in any [stablecoin/token I believe in] pair.

  21. It seems to me one, of the pair, goes up to even the amount. On the platform I’m using… That’s just what it looks like to me. Am I incorrect? I’d add a picture but I’m a little wary of doing that..

  22. It’s not necessarily loss. It is reduced volatility. Which goes both ways. In a bull market you underperform the underlying and in a bear market you overperform.

  23. It’s actually a loss, and that’s why I commend Alliance Block for working on a DEX with a built in AMM that solves this.


Leave a Comment