Get 777 Sense of Shapes NFT for the most prioritized access to ecosystem.

nftperp 204 days, 9 hours ago

Look At My JPEG — Ep.29 Recap — Upshot

Welcome to another summary of our recent conversations with NFT Finance teams, where we delve into their projects, market views, future plans, and much more. This time, we had the pleasure of speaking with Nick Emmons from Upshot, a leading NFT appraisal platform.

Introducing Nick Emmons

Nick Emmons has a long-standing history with cryptocurrency, tracing his involvement back to around 2015. Before founding Upshot, he led the crypto department at John Hancock and Manulife, a large asset management and insurance company. His work focused on pricing esoteric assets, with a particular emphasis on risk assessment, and researching decentralized systems for non-parametric risk mitigation and insurance. When asked about his first NFT purchase, Nick mentioned that it was probably a CryptoKitty, purchased at the end of 2017. CryptoKitties, as noted in the podcast, was one of the first Ethereum applications to highlight the scalability issues of the blockchain, sparking discussions and subsequent projects focused on Layer 2 solutions.

About Upshot

Upshot emerged from his previous work in the insurance sector at John Hancock, particularly around developing decentralized systems for risk mitigation. They initially considered using their crowd-sourced technology for insurance but quickly pivoted to the NFT space. According to Nick, NFTs, like real estate or physical art, are “non-fungible, low-velocity assets,” meaning they aren’t traded frequently. This characteristic presents a unique challenge for valuation, making an appraisal mechanism crucial.

Upshot’s first iteration employed a crowd-sourced approach for NFT appraisal, similar to a “hot or not” concept where users would compare two NFTs and weigh in on their value. However, they quickly realized that this model was neither scalable nor particularly accurate. Hence, they pivoted to machine learning (ML)-powered appraisals.

Over the last two years, Upshot has been fine-tuning its ML models for NFT pricing, and according to Nick, the models are now robust enough to price over 87,000 collections of NFTs on a sub-hourly cadence. This shift towards ML-powered appraisals is part of a broader strategy to open up new possibilities for financial products centered around NFTs. Nick believes that by providing a reliable price feed for NFTs, they can bridge the gap between decentralized finance (DeFi) and the NFT world, thereby unlocking a whole new “design space for interesting financial primitives.” The platform has not only iterated on its core appraisal model but has also expanded horizontally to appraise individual NFTs, and collections, and even index a number of collections of NFT evaluations.

Upshot’s Bread and Butter — API

One of the key applications of Upshot’s API is in NFT marketplaces, where it offers a means of estimating the “true” value of an NFT beyond the basic floor prices. It allows users to identify potentially undervalued or overvalued assets in real time. Additionally, the API’s data is becoming “oracleized” to be more consumable on-chain, directly influencing DeFi protocols, lending protocols, and automated market makers.

Addressing the question of how the API determines NFT value, Emmons described a multilayered approach that focuses on traits and market activities. The algorithms take into account the combination of traits an NFT has, comparing them with other NFTs in a given collection to deduce their relative worth. Traits can include explicit characteristics like color and texture but also include more abstract elements, like “cleanliness” or “cohesiveness” of design elements. Social sentiment, community interaction, and transaction history are other factors that feed into the model.

The challenge of appraising NFTs, especially those that are not frequently traded (termed “low-velocity”), is in providing accurate valuations. Upshot continually assesses the accuracy of its model by comparing its appraisals with actual market sales, scrutinizing margins of error to improve its algorithm. The model also focuses on enhancing its accuracy for the ‘tail’ of collections — i.e. Grails like Zombie Punks or Spirit Azukis.

Emmons also touched upon the concept of “meta traits,” traits that aren’t explicitly defined but have an influential role in an NFT’s value. These meta traits can sometimes require a human touch for identification, as they may encompass cultural or memetic aspects that aren’t easily quantified. In the future, community-sourced input may help identify these more effectively.


In this segment of the podcast, Nick Emmons discussed Upshot’s aggregator feature, which allows users to connect their wallets and trade NFTs across multiple marketplaces. This feature also integrates Upshot’s core appraisal system, providing real-time valuations on the screen. Emmons stated that the aggregator was initially designed as a proof of concept to showcase how appraisals could be incorporated into various market flows or spot market tools. It’s also the company’s first Business-to-Consumer (B-to-C) product focused on these appraisals.

Astaria and NFT lending

The discussion delved into the partnership between Upshot and Astaria, emphasizing how Upshot’s appraisal API and tools aid in the NFT lending and borrowing process. Upshot aims to bring a more nuanced valuation system to lending by appraising each NFT’s actual value rather than simply considering floor prices. Nick explains that Upshot serves as a strategist in the Astaria system, which takes an innovative approach by decoupling lenders and capital providers. This allows for a more automated and tailored lending process by using constantly updating terms based on individualized appraisals and predicted future asset values.

Astaria’s system allows borrowers to autonomously accept loan terms against capital deposited in a vault. Upshot’s role as a strategist involves providing individualized loan terms for specific NFTs based on appraisals and future value projections. Their system also factors in broader market conditions and the vault’s portfolio construction to mitigate risk and optimize capital efficiency. The ultimate goal is to give borrowers a personalized, efficient loan experience while ensuring that lenders can accurately assess risk and set proper terms. Emmons elaborates that the most challenging part of NFT lending is predicting the future value of an asset to generate proper loan terms, likening the activity more to options trading than traditional lending. Upshot’s integration into Astaria aims to solve this by utilizing its robust appraisal mechanism to forecast an NFT’s future value. This predictive element is key to determining the loan-to-value ratio, interest rates, and other terms of a loan. The discussion also highlights the importance of individualized loans in NFT lending, emphasizing that generic approaches leave a lot of capital on the table. Unlike traditional assets, NFTs are non-fungible, making each unique and requiring a more nuanced appraisal method. Emmons states that at launch, they plan to support eleven different collections, offering terms for every NFT within those collections.


The discussion touches on Upshot’s GMI score, an intriguing metric integrated into the platform’s front end. The GMI score serves to evaluate the overall success of NFT wallets or collectors and plays a significant role in the asset valuation process. Specifically, the score aids in understanding the quality or “caliber” of the holders of particular NFT collections. The rationale behind this is that the value of an NFT can be significantly influenced by its holder base, thereby impacting the valuation of individual assets within that collection. In this context, GMI is far more than just a “gimmick” or “fun thing.” It becomes an essential feature in their pricing model and holds importance in wallet-level or community-level analytics. Given that much of the NFT space is an experiment in online community building, the “caliber” of a community can be an essential determinant in projecting whether the price of a collection is likely to rise or fall.


The final segment of the conversation delved into the topic of NFT indexes, a recent innovation by Upshot. The indexes serve as an important tool for tracking different aspects of the NFT market and offer an avenue for investors to gain exposure to the space. Unlike individual NFT collections, these indexes abstract away the need for active management of a portfolio, thereby making it easier for people to invest in the NFT market as a whole. One significant potential application is the integration with synthetic financial instruments like perpetuals, which could offer users synthetic exposure to different NFT baskets. This would enlarge the kinds of capital that could be onboarded into the NFT space.

However, the conversation also touched on challenges like liquidity and the centralization inherent in managing these indexes. Rebalancing the index — changing asset weights, adding new assets, and removing others — poses a question of decentralization and trust in the algorithms governing the index. Upshot is working on mitigating these concerns by exploring ZK-ML (Zero-Knowledge Machine Learning) and other decentralized technologies to bring verifiability to their models. The goal is to ensure that as these markets grow and evolve, the data is generated trustlessly and is easily verifiable.

Big thank you to Nick for joining us and having a thrilling discussion! Follow him on Twitter:

Listen to the full podcast episode with the following links:

Twitter Spaces Recording (limited time only!):











Look At My JPEG on Spotify:




Crypto Punks


Other Articles