Relative Performance Token (RPT)

Hello All,

I have an idea for a financial asset that I believe would be particularly relevant within the crypto space. It can be used as an leveraged vehicle for investors or as a tool to manage treasury. I am posting my idea here because I would love to hear all of your opinions and I would like to know whether it’s possible to build with UMA. I would love to team up with dedicated developers to make this happen.

Any crypto investor can relate to the intense volatility in this space. I love leverage, but have been destroyed by it several times - mainly because of unlucky timing. Any FUD event can completely liquidate your investment, even if your belief in a particular technology is strong. Furthermore, BTC and ETH define the market and move altcoins in a particular direction - whether that is warranted or not.

My idea is to build a leveraged financial asset that measures a specific token’s performance relative to a basket of other tokens - a relative performance token (RPT). For example, I am an UMA-maxi and I believe that it will always outperform its peers, whether that be in a bear or bull market. I seek leverage, but I want to be protected from absolute macro events (i.e., FUD). I can design an asset that measures UMA’s token price over a specified time frame (i.e., return) (let’s say, one year), relative to a peer group of other tokens. That peer group can be anything really - Defi assets, Top 100 tokens by market cap - but probably should be something relevant. At the end of the one-year measurement period, each asset’s one-year return will be measured, and the group of tokens will be sorted from high to low return. The payout of the RPT will be based on UMA’s ranking within that group (percentile rank). It could be binary: payout of 2x for upper quartile performance / 0.5x for lower quartile performance / 1x for median-ish performance. Or, it could be linear along any sort of payout curve based on the measurement asset’s percentile rank. You can of course create an RPT for any token (Sushi, Uniswap, etc).

The pricing of the RPT token will likely need to be contingent on several factors - the asset you are measuring, the peer group, historic volatility of the assets, the payout curve, etc. - but I believe it is possible with some financial models, particularly a Monte Carlo simulation. There are, of course, kinks to this design that will need to be worked through - like the price of the RPT, whether its price fluctuates with demand, and the details with how the payout works, all of which I think are achievable with the right minds at work.

A little bit about myself - I work at an executive compensation consulting firm. We are engaged by the board of directors of major corporations to design incentives (primarily performance-based incentives) for C-Suite executives. Although I love my work, my hobby has been crypto over the past couple of years. I love discovering new protocols and learning about new financial applications. Part of my work is designing incentives for CEOs, CFOs, and other C-Suite folks. The vast majority of their compensation is equity-based (~50-70%), and about 50-75% of that equity compensation is performance-based equity or performance share units (PSUs). Executives receive target shares that ultimately payout at the end of a given performance period (typically three years) based on a pre-determined performance measures. Companies can use financial measures like absolute revenue growth, EBITDA, or ROIC. Or, they can choose to measure performance on a relative basis. I have experience designing relative total shareholder return (rTSR) performance share units for these executives - which measure a company’s stock price return relative to other companies. Essentially, these function similarly to the design I noted above. Executives receive target shares (let’s say 100K), that may increase for upper quartile stock price return versus a peer group (200K shares), or decrease for below median performance (50K shares, or even 0). This design protects the executive from absolute market movement, while providing leverage both in the amount of shares that the executive earns AND the stock price appreciation. Relative measurement is important to retain the executive during volatile macroeconomic events. For example, during the recent pandemic. If I am an executive of a major restaurant, I likely would not hit my financial performance targets during the pandemic, however, I could earn my relative TSR PSUs because I may outperform other restaurant peers (my company may be more financially stable, or have a digital platform, that the market will price into my stock).

I think this type of design can be applied to the RPT, and would be very helpful for investors who seek leverage, yet want to be protected from macroeconomic events affecting all coins. Investors may even using the RPT as a hedge. And, protocols can use this vehicle to diversify their treasury while staying loyal to their native token.

I do not have a developer background - but I believe my experience designing these assets for my clients is invaluable and - when combined with a team of dedicated developers - we can bring this asset to life.

I welcome all thoughts on the idea!


Looks interesting, and it should defo be possible to build this on UMA, but possibly quite complex.I wonder if it could be simplified as an initial exploration by using existing indexes - such as DPI or DEFI5 as a base measure.

It could be particularly useful as a particular form of KPI option which tracks token performance relative to < chosen subset of market >

1 Like

Thank you Mhairi! Great point. Simplifying the asset using a KPI option makes sense - especially as a starting point given these indices are already developed.

One difficulty I foresee with using an index is that you would have to pre-determine the return goal that would yield a certain payout. For example, DPI Return + X% = 2x payout. Defining that X% can be tricky in a volatile market, but can likely be done by looking at historic returns of an index. When using constituents, the performance goals are implied (75th percentile = 2x payout).

Really good idea @Cryptowsky! I’m the co-founder of and I’d love to work with you on this concept. It’s somewhat complex but completely doable.

Let’s talk.

Over on our discord, @bitznbrewz was mulling this over using comparative rankings with the LSP project. I dont want to misrepresent his thinking so will leave it for him to explain

@kevin thought you might also be interested in this discussion.

1 Like

I think we could use the KPI options and the LSP. For example, IndexCoop could put up $X of DPI as collateral and mint long & short tokens. They could distribute the long tokens to members of their DAO and sell off the short tokens. Or, for full chaos, give them to their competitors.

As an example metric, we check the year-to-year performance of DPI versus sDEFI, DEFI5, and DEFI+L. We measure the performance, relative to ETH, for each of these tokens. The tokens are ranked by their one-year performance with rank 1 being the best performing and rank 4 the worst.
For the sake of this example, I measured the performance against ETH since DEFI tokens tend to have strong correlation with ETH instead of USD. Instructions are provided to voters on how to determine the rank of DPI against these other indices. All of this semi-complex computation would be done by voters off chain, since these tokens cannot be liquidated.

The following values would be passed into expiryPrice in the contract

Rank 1: 1
Rank 2: .66
Rank 3: .33
Rank 4: 0

This ratio represents the collateral payout to the holders of the long tokens. The short token holders receive the inverse.

I’m still working with some other members of the UMA team to get the technicals sorted. I just wanted to chime in on here because I saw your Twitter thread on this topic today.

@bitznbrewz thanks for sharing your thoughts! I’m glad you found this post interesting and that you have already started thinking through a design. Combining KPI options and the LSP makes sense to me. I do have a few comments for your consideration:

  1. Do we need to measure performance relative to ETH? That may be overly complicated because you can simply measure one-year return for each token. When you evaluate the tokens at the end of the measurement period, the return that is driven by/correlated with ETH washes out because we assume that is baked into each token’s return figure.

  2. I think using tokens would be more helpful than measuring performance of indices. By evaluating DPI versus other DeFi indices, you would be measuring the efficacy of index construction. Many of the indices will have the same constituents. Some may differ by one or two constituents or even the weighting of the constituent.

I think measuring using index constituents or a set of tokens would suffice as long as they’re relevant. See my example below – which did not end up in my twitter thread lol :slight_smile: . These are the current constituents of DPI. If this RPT is designed to measure Sushi’s relative performance from 6/20 to 7/20, then Sushi would score at the 42nd percentile (or ranked 8th per your methodology).


  1. Instead of determining the payout KPI options based off of the rank of each observation, can we design it based off of a few goals and linearly interpolate between them? I imagine the more KPIs you have the more technically complicated it is to build (although I could be wrong). By linearly interpolating, you can have fewer KPIs, but still have a large measurement group. Example KPIs: below 25th percentile = 0% payout, 25th percentile = 50% payout, 50th percentile = 100% payout, 75th percentile = 200% payout. Then, you can create a simple function that linearly interpolates payouts for percentiles between these.

Using this payout function and my example above, this Sushi investor would probably realize a 84% payout. So, if they invested $100K in this RPT it would be worth $84K at the end of the period. Had they invested in an option, it would be worthless. A token would be worth ~$75K.

  1. Final question, which may be too soon, but how do you envision pricing this asset? I have a few ideas, but would love to hear your thoughts.

Happy to work together on this project. Feel free to reach out on twitter or discord if its easier.

In the forgotten dustbin of tradfi, alpha options once almost existed (briefly traded a few lots) back in 2011ish. Here’s the paper by the inventor of the VIX no less. Trading Relative Performance with Alpha Indexes by Jacob S. Sagi, Robert E. Whaley :: SSRN

Highly recommended reading. I attended one of the launch discussions at the NYU volatility institute back in the day, it’s a fabulous product of you can get it to work.