# Why APY Needs To Die (At Least For Now...)
*13 min read*
> "[We] were spending much of each day trudging to each other’s offices to complain about the crazy deals – characterized by low returns, high risk for investors, and a lot of optionality for issuers – that were easily being brought to market. “If deals like this can get done,” we agreed, “there’s something wrong with the market.” Few people, we thought, were demonstrating prudence, discipline, value consciousness, or the ability to resist the fear of missing out. **Investors are supposed to act as disciplinarians, preventing undeserving securities from being issued, but in those days, they weren’t performing that function. This signaled a worrisome state of affairs.**"
> - Howard Marks - Taking The Temperature July 10, 2023[$^{[1]}$](#get_1)

What makes an investment investable? Or better yet, given a plethora of investments, how might you go about choosing which is *best*? There are many ways to analyze a potential investment - fundamentals, technicals, gut instincts - but almost all of them are ultimately guided by the same thing: historical performance. Evaluating the historical performance of an asset is one of the most common ways that we assess its quality and set future expectations. But the performance metrics used to gauge historical performance in traditional finance aren’t being used correctly by the cryptocurrency industry, which warrants reflection about their continued usage.

### TL;DR
- Performance metrics are an important tool for evaluating investment decisions. Some are better than others.
- APR and APY can be misleading when applied to cryptocurrency markets, potentially doing more harm than good for potential investors.
- Modifying these metrics to account for factors specific to cryptocurrency and decentralized finance can make them more applicable and provide a clearer understanding of an investment and its potential return.
- To take the next step on the path towards legitimacy and mainstream adoptions Decentralized Finance (DeFi) needs to develop its own standards for data reporting and communication. These standards should be part of self-regulation that other projects, protocols and teams adhere to without any external intervention.

### Background
Reporting metrics for investment performance are useful tools for investors making decisions. Understanding a rate of return or yield is an easy way to compare performance across assets and products. It doesn't say anything about risk, but it is commonly understood that the potential for greater returns implies greater risks. This is known as a risk to reward ratio. Now, not everyone sits down to quantify "risks" as risks is a rather amorphous term which includes things such as "acts of god" and "electro-magnetic pulses" as well as "exchange insolvency." Some of these risks can be dealt with through the use of risk mitigation strategies, others are left in the hands of the universe.

Maybe this post has already lost you; the point is: Surfacing information is essential for investors to be able to make informed decisions and set realistic expectations. This is not a particularly revolutionary or even controversial thought. Beginning in the early 20th century in the United States, as consumer lending began to accelerate[$^{[2]}$](#get_2), APR became a ubiquitous metric used for lenders and consumers to better understand expectations and provide comparisons. APR remains a standard metric throughout the borrowing and lending industry to this day. If you've ever interfaced with the banking or credit system, you may have noticed that your interest rate, whether that be money owed on a loan or money earned in a savings account, is denoted in APR.


Annual Percentage Rate (APR) is a metric used when evaluating the expected return in excess of your principle (or owed). This return can be generated through a number of means, for example on a savings account, that APR comes from the bank lending out that money at a higher rate than what is paid back to you. Or in crypto, sometimes it's emissions of a token as a "reward."[3] APR alone is a powerful metric to help guide estimations for return, however what happens when you add your returns to your principal? Like a Solana Stake Account, this is where Annual Percentage Yield (APY) comes into play. The key is one takes into account compounding your returns. So you may have a 6% APR which would roughly amount to 6.12% APY.

However, not all metrics are calculated the same, and businesses may have drastically different ways of applying the math or what might influence such metrics (we won't even touch on IRR[$^{[4]}$](#get_4)). Within developed systems, these calculations and displays are governed by representative bodies ensuring that they meet standards and expectations. In the U.S. in 1968 the "Truth in Lending Act" (Regulation Z)[$^{[5]}$](#get_5) was enacted which was implemented by the Consumer Financial Protection Bureau and funny enough how we ended up with "zero percent APR" auto loans.

Crypto can stand on the shoulders of giants by adopting reasonable standards for surfacing data and reporting historic investment performance. Understanding why things are in place is critical for the choice of adoption and re-inventing finance doesn't demand we throw the baby out with the bathwater. There are reasons (often really good reasons) things are done the way they are, and choosing to ignore history is never a path towards success.

### Why do metrics matter?
The way that most people understand and interpret APY is through the lens of traditional finance. But this interpretation doesn’t neatly apply to decentralized finance, where highly volatile markets with non-traditional structures are the norm. In fact, using a traditional understanding of APY can detract from an investor’s ability to make informed decisions.

*Katie Kerpel © Investopedia, 2019.*

The ability to calculate and display APY is simple. It is not challenging to implement the formula for integration within display charts and UI components.

$$A = P*(1 + {r \over n} )^{(nt)}$$

And to be fair, these displays are accurate in the singular result of X token yields X+Y%. The issue is the advertised APY and the resulting graphs and charts assume nothing changes and the investment is given adequate time to compound. That, however, is not the way performance is realized. Things in crypto do change…fast. Investment half lives are measured in days and weeks. Price fluctuations and their resultant losses can be realized in seconds (see toxic flow[$^{[6]}$](#get_6) and markout[$^{[7]}$](#get_7)).

For example, a 20% APY isn’t a meaningful metric for assessing an investment if the token is inflating at 50% per year. It is doing more harm than good. The investor needs to evaluate beyond the APY to fully understand the expected return but often will not, as they assume that their traditional understanding of APY still applies.

The return doesn’t come close to matching the APY in notional value or token denominated yield. APY is a misnomer in this case

Historically, yields which are (or were) commonplace in crypto would be indicators of Ponzi schemes, fraud and con-artists in the traditional market space. Because of the dynamic nature of cryptocurrencies and competition for liquidity, the APYs seen throughout DeFi are often much larger than anything seen in the world of traditional finance. It’s not to say these metrics cannot be used, but we need to heavily qualify or come up with a new term to ensure the expectation from investors are built on a more complete understanding of the realities of DeFi.

### How did we get here?
When you deposit an asset into a yield bearing instrument such as a savings account, staking account or a peer-to-peer lending platform, the APY is an extremely useful metric. Compounding returns are a powerful effect that’s crucially important to understand when evaluating a potential investment. So borrowing APY from traditional finance and applying it to DeFi strategies was a no-brainer. It tells the investor that if you deposit X, you should expect to receive yield Y.

That, however, isn’t the end of the story, because where this works well is currencies and assets denominated in said currency — more specifically state sponsored fiat currency— doesn’t always work well in complex financial instruments like we see in DeFi.

Drilling down incredibly complex interactions of DeFi systems and their resulting performance is a massive challenge. Especially so when you consider that some protocols are simply a layer ON TOP of underlying tokenomics and other protocols. The composable nature of these systems cannot be neglected when considering the emergent challenge to convey whatever metrics are presented to a user. By the time a token makes it into a liquidity pool or is being staked, there can be many associations or implications which encumber this asset BEFORE it’s integrated within another system.


For example, a Liquid Staking Token (LST) earns yield from staking. But when you put that token into a liquidity pool, the expectation for yield (as denoted by its advertised APY) depends on what it is paired with. In the case of an LST/SOL pair, the LST should always accrue value relative to the Sol it is paired with. This detail is obscured by the APY, which only accounts for the trading volume of the pair.

Because of this complexity it is not fair to say that most or all parties who are displaying a single-input APY without fully accounting for all of the potential sources of yield and loss are doing it to deliberately mislead investors.

Within the crypto industry the marketing of products is very competitive. This competition is driven by attention, liquidity and ease of use. Some of these drivers beget others, but if a competitor is willing to accept more risk and investors are willing to invest with the competitor, it’s a race to the top with risk acceptance (until an inevitable collapse) and marketing techniques and strategies. Therefore once one adopted the use of APY, others followed suit. It’s a relevant metric after all, and if you want to entice liquidity with incentives, what better way than big numba?

### Why are things broken?
APY doesn’t give crypto investors the full picture but, due to the pervasive traditional understanding of APY as a performance metric, many investors don’t realize this and aren’t inclined to dig deeper. Capital is being misallocated where the expected returns are actually not even remotely close to the realized returns.

In some cases, the discrepancy between advertised APY and realized returns are waved away with a reference to Impermanent Loss (IL) as if it’s some magic wand, Schrödinger’s cat metric whereby if you just hold out long enough (or, sigh, time your entry/exit), everything will be as promised.


Impermanent Loss is a concept termed by Liquidity Pools to help explain the results of keeping your investment in a LP. When you make an initial investment you will be awarded a certain amount of tokenized representation of the pool, a % based on your contribution over the total capital in the pool. These tokens represent a claim for redemption or receipt for subscription. However in a LP over time the ratio of base to quote tokens changes given many LPs depend on a Constant Function Market Making implementation, x = ky or for those who want something they should remember from grade school y = mx+b (with b, as 0). Yep the equation for a line[$^{[8]}$](#get_8).

So you will get your base (rise) over quote (run) at particular times depending on the activities of the pool. These activities are trades (a.k.a. swaps) which affect the configuration of the pool and adjust the ratio. Now IL implies that if you were to redeem your LP tokens for their respective configuration (base/quote) you may have unrealized losses due to the change in pool make up. But those losses aren’t realized, unless you realize them, and may not last forever. There are a number of issues with IL and LPs which we won’t address in this piece, but I think it’s important to consider this metric in conjunction with APY as it’s a major factor for where we see APY so prominently displayed (why don’t we prominently display the average IL too?!).

Therein lies the problem, the outcome of investors continuing to invest in these products with poor performance is absurd, when faced with the stark contrast reality. There are better uses and better products they should be investing in to further the development of this industry. Instead they’ve been captured and fed false promises, and in some cases, kept away from the data that proves otherwise.

At best it’s misinformation, at worst it is fraud. If we assume Hanlon’s razor here that it’s just incompetence of the parties displaying this information then the problem is just that. If at worst, these parties realize that the resulting product is losing money for its investors and they are deliberately obscuring this result, then they need the social pressure to be revealed to investors to allow for capital to become informed and move (even at the detriment of the current status quo within the industry).

Here in the future of finance, we should demand better not worse than the counterpart. We can’t get away with things because there is no oversight. We need to ensure, fundamentally, these are core components of the protocols pushing us ahead. We don’t get better with hacks and exploits, we don’t get better when it’s easy to scam people, these are detrimental experiences which reinforce the status quo, not extend the capacity of those willing to think beyond it.

The inability of this industry to take itself seriously is a massive detriment and not only further isolating itself from the potential brain trust willing to accept risk, but also from an industry it will need to cater and on-board to facilitate the use of these primitives for humanity. We suffer together, in all the bad news, and only grow as fast as those suffering the most.

### How can we do better?
**Get rid of APY for the time being** - It’s a useless metric in this industry without HEAVY caveats or context. Therefore if we do want simplified metrics I’d recommend we have components constructed around tokenomics, protocol, risk and external parameters.

**Discounted APY** - For one example a Tokenomics Adjusted APY (TAAPY) would be a great first step, where emission schedules with trading volume and historical pricing is taken into account when projecting a return. For another, if an insurance fund has the capacity to refund 80% of user funds in the event of an exploit, we should be discounting APY if that’s a component. There are a number of SOUND practices which could be established to help incentivize better design and marketing, such that the investor expectation and the reality experienced are not so disjoined.


**Transparency** - It’s blockchain after all, this information should be readily available and teams and businesses need to invest in surfacing this information, regardless of the good or bad news it may reveal. Dune/Flipboard dashboards are a minimum effort, what would be better is if we had a team of external parties validating and verifying the results from some established criteria. Think like Neodyme, but for accounting and reporting.

**Reporting** - The current traditional systems have reporting structures, so should we. This isn’t something hard to nail down or to generate at any time, the data should be easily surfaceable and accessible, with these things becoming mundane and expected. I know I’ve seen some products and services begin to add more qualifying metrics to their data display, but I think we can do more with something like quarterly reports which everyone links on their sites and its as expected as having a X (formerly known as Twitter) handle and Github. Things like average return for a $1000 investment across X instruments / products, average user return compared to protocol revenues just to name a couple.

**Review** - The community should make a push to establish funding for time and resources in order to have each product/service/protocol be reviewed. These things must be taken under consideration and probably best under purview of those in competing markets.

**Messaging** - These reports, reviews, dashboards and information needs to be a crucial component of any brand and marketing messaging put forth by a team and business. There should be an established cadence where the social layer expects these revelations and anyone not participating can be evaluated as a bad actor.

**Social Norms** - Even if the resulting information is not well received, the social layer should accept the fact that these things are being revealed at all, without the intervention of a state actor. There needs to be a balance in these systems that encourages transparency regardless of the outcome. As crypto doesn’t have a lender of last resort, collapses will occur, it’s better to have the flight occur because a protocol is known to be a failure, than it is to blindside users.

**Self Governance** - Any step forward is a proven attempt that we have the capacity to enact change for the better. Any attempt coming from the industry that knows itself best, would bear the best outcome for those who interface with it. If we can ourselves come up with standards and practices even without the capacity to enforce, it would be a step for others to follow.

### Conclusion
The way APY is used in its current form within the cryptocurrency landscape is of detriment at best. We should not encourage its use and should develop our own metric to help better inform investors. We understand that APY is easy to calculate and that the composable nature of these assets and products make a more appropriate calculation incredibly complex, but we must strive to do more to describe risk and return. As new information is surfaced and results of performance are known, I believe full transparency and data display are paramount to the success of this industry.

Investors are being misled (either deliberately or not) and their expectations are not matching the outcomes. We can and should do better by adopting our own standards and practices such that the competitive landscape is more reasonable and prepared for a liberating financial future.


### References

<span id="get_1">[1]</span> Taking The Temperature. Howard Marks. July 10, 2023. https://www.oaktreecapital.com/insights/memo/taking-the-temperature
<span id="get_2">[2]</span> The Long History of “Truth in Lending.” Anne Fleming. 2018. https://scholarship.law.georgetown.edu/cgi/viewcontent.cgi?article=3088&context=facpub
<span id="get_3">[3]</span> Cosmos community approves capping Atom inflation rate at 10%. Zack Abrams. November 25, 2023. https://www.theblock.co/post/264683/cosmos-community-approves-capping-atom-inflation-rate-at-10
<span id="get_4">[4]</span> Internal rate of return: A cautionary tale. John C. Kelleher, Justin J. MacCormack. August 1, 2004. https://www.mckinsey.com/capabilities/strategy-and-corporate-finance/our-insights/internal-rate-of-return-a-cautionary-tale
<span id="get_5">[5]</span> Truth in Lending (Summary). Office of the Comptroller of Currency. https://www.occ.treas.gov/topics/consumers-and-communities/consumer-protection/truth-in-lending/index-truth-in-lending.html
<span id="get_6">[6]</span> Follow-Up Analyses of LP Profitability in Uniswap V3. CrocSwap. January 6, 2023. https://crocswap.medium.com/follow-up-analyses-of-lp-profitability-in-uniswap-v3-2cfc8c5e014e
<span id="get_7">[7]</span> Solana DeFi Markouts. JarryX. October 2023. https://dune.com/jarryx/solana-defi-markouts
<span id="get_8">[8]</span> Impermanent Loss Explained. Binance. May 18, 2023. https://academy.binance.com/en/articles/impermanent-loss-explained