## Intro

Upshot is entering the NFT lending space. By leveraging our appraisal models and new infrastructure built specifically for lending, Upshot will be launching managed NFT lending strategies. Our lending strategies use different methodologies that can appeal to a range of liquidity providers with different risk profiles. In this post, we’ll outline how our appraisals enable us to construct efficient portfolios of NFT loans, details into the new lending-specific models that we’ve built over the past couple of months, and how we’ll be bringing these lending models to market in the coming weeks.

## What Goes Into a Successful Lending Strategy

When constructing terms for NFT loans, it’s vital to understand the value of the NFTs you are lending against. The amount you loan against an NFT relative to its value is known as the “loan-to-value” (LTV). As a lender, if your LTV is too high, you are exposing yourself and your backers to undue principal risk. On the other hand, if your LTV is too low, you are leaving money on the table and will likely lose borrower interest to competing lenders.

This is where appraisals fit in. NFT appraisals give us a reliable in-the-moment estimate of the value of an NFT, allowing us to set loan terms that are competitive while mitigating our risk.

### Appraisals

Upshot's appraisals are among the most accurate and robust models for NFTs on the market. Every hour, over 74 Million NFTs from more than 67,000 collections are appraised by our ML models with industry leading accuracy (MAPE of 3-10% for our best-performing collections). It is these appraisals that are powering our lending strategies.

While appraisals are at the core of our lending strategies, we have built several other components that are essential to being a successful lender in the NFT space. These other pieces include forward-looking estimates of asset value changes (recovery rates), default probability estimates, and our risk management framework. A quick description of each follows:

#### Recovery Rates

Our recovery rate model calculates forward-looking estimates of NFTs’ value, which takes into account the current market situation and helps the lending strategy to predict the future collateralization of an NFT-backed loan. Without recovery rate estimates vs the current fair value (our appraisal) a strategist cannot estimate how much money they stand to lose when a borrower defaults.

#### Default Probability Model

Like TradFi credit providers, our models factor in the probability of a borrower defaulting on their loan. This model helps us to estimate the possible frequency and likelihood that the borrower misses an interest payment and assess the chances that we will have to liquidate the assets that we are left in possession of.

#### Risk Management Framework

Our risk management framework is where our concentration analyses occur and where we manage the number of collections we are writing loans against. This framework also takes into account the loan concentration per collection on other major lending platforms, ownership concentration per collection, our own loan concentration per collection, and collection concentration per wallet. By not only acting at an individual NFT level, but also at a holistic collection and market level, we can protect both liquidity providers and ourselves from any potential downsides.

The number of collections we lend for is also not fixed. When market conditions improve we are able to provide terms for more collections, but when there is volatility in the market our framework automatically scales down the number of collections we lend for.

## Striving for Breadth

Because of Upshot’s appraisal coverage, we are capable of offering relatively safe and competitive terms for dozens of collections. This is achieved by adjusting the collections we publish terms for based on our fair value estimates at the horizons we lend for. To decide which collections are suitable, we analyze the volume, liquidity, wash trading as a percentage of total trading, ownership and transaction wallet concentration.

The criteria that go into selecting collections are volume, liquidity, wash trading as a percentage of total trading, ownership and transaction wallet concentration, over multiple months, so that we lend for more stable but also diverse collections that do not suffer from and are less prone to manipulation.

## What’s Ahead

With this lending infrastructure and strategies in place, we will start to bring them online in the coming weeks. We will be underwriting loans directly on leading NFT lending platforms, but we will also be one of the initial strategists on Astaria when they launch. As an Astaria strategist, it will be trivial for anyone to get exposure to our strategies’ performance through the Astaria protocol. We will be releasing some more information about how our strategies work, the types of loan we’ll be underwriting, and more in the next few weeks.

Upshot has entered the NFT Lending space. Our lending strategies, which leverage our appraisal models and new infrastructure built specifically for lending, appeals to a range of liquidity providers with different risk profiles.

Introducing: Upshot Lending Strategies


At Upshot, we believe a robust interpretation layer for NFT data is key to the success and growth of the NFT industry. Keeping this in mind, we built collection recommendations: A feature that enables users to better discover NFT collections based on their current interests and holdings.

This article will cover why collection recommendations are important for the NFT space as a whole, how we used AI to put personalized curation front and center for our users, and of course, how you can get started using them today.  

[Try it: Collection Recommendations is live on our aggregator](https://app.upshot.xyz "Upshot App")
![Upshot Web App with Collection Recommendations feature](//images.ctfassets.net/lhkcghnr8ql4/0ZW23Boca6kirS9hdq9W4/030b3ebbf9f4fee19b55506cbfbe6aa7/Screen_Shot_2023-02-28_at_6.53.57_AM.png)

## Why Recommend

The NFT space is exploding. Amongst many things, this has led to an ever-increasing number of collections in various categories such as PFPs, gaming, music, generative art, etc.

With more than 160K collections (at the time of writing) available through various marketplaces, it has become increasingly more challenging for individuals to decide which collection to focus on. 

Analogous to the evolution of the e-commerce industry, the NFT space will soon realize the importance of content curation and personalization for individual wallets. To create this will need to develop data-driven techniques that reduce the search space for wallets and help individuals navigate the sheer volume of content produced in the NFT space.

### For the User
A recommendation system uses algorithms that take a large set of items as input to determine what to display to a user next based on those inputs. Recommendations are part of our everyday life - think of Facebook News Feed, YouTube's homepage, and Netflix's "Recommended for you" navigation page.  At Upshot, our recommendation algorithm can infer “what the user may like to own” based on the collections they currently hold in their wallet. 

### For Developers
Using our collection recommendations algorithm (which is bundled into the [Upshot API](https://docs.upshot.xyz/docs)) developers can create a better, more personalized experience for their end users. 

This personalization can lead to new products, enhanced user stickiness and transaction volume, and an overall better experience in the NFT space - one that cuts through the noise and presents the relevant information on a specific per-user basis.

## Our Methodology

We based our initial approach on inferring the wallet's preferences for collections based on what the wallet already owns. 

For baseline recommendations, collaborative filtering is the first step. This step doesn't require too much information about collections except for which collections specific wallets have traded, owned, viewed, etc. 

Collaborative filtering analyzes the relationships between wallets and interdependencies among collections to identify new wallet-collection associations. The technique recommends collections to wallets based on the preferences of other wallets by identifying similar wallets or collections. 

In our case, we use NFT ownership as a wallet's preference. A critical step in the recommendation process for a given wallet is the identification of similar wallets (known as user-based) or similar collections (known as item-based).

At Upshot, our recommendation technique is item-based where similarities between collections are leveraged to determine the associations between wallets and collections. This item-based technique is explained using Figure 1 which shows wallets ownerships of collections.

![collaborative filtering figure](//images.ctfassets.net/lhkcghnr8ql4/7zDQO0delNgh4CI49G11Qi/c79f69078e745358f228d44b0b5b2bdb/collaborative-filtering.png)
Figure 1: Item-based Collaborative Filtering

To recommend more collections to wallet 1, we first find the collections owned by wallet 1 which are collection 1 and collection 2. Now, the next step is to find similar collections to collection 1 and collection 2. As collection 2 is also owned by wallet 3 who also owns collection 3 and it is not owned by wallet 1, the process will present collection 3 to wallet 1 as one of the recommendations.

From a technical perspective, the similarity between collections is determined using Cosine Similarity, which calculates the similarity between two vectors by calculating the cosine angle between them. The advantage of using this method is that it is invariant to vector size as in our dataset, the number of collections each wallet interacts with varies a lot. This difference in magnitude does not affect the similarity measure since it only considers the angle between vectors instead of their sizes. 

Mathematically, cosine similarity is calculated as the dot-product of collection pairs divided by the product of normalized vectors. i.e.

$$
cos(\theta) = \frac{\sum_iA_iB_i}{\sqrt{\sum_iA_i^2} \sqrt{\sum_iB_i^2}}
$$

Once a collection-collection similarity matrix is available, the next step is to calculate a score between the wallet and every collection not currently owned by the wallet. The association between a wallet and a collection is the weighted average of a set of collections that are similar to collections owned by the wallet. Notice we use weighted average where weights are the similarity score between collections. The weighted average reduces the bias in situations where similar collections are not equally similar to the owned collection.

The final step in the recommendation process is ranking the wallet-collection associations in the descending order and determining the top 5 recommendations per wallet. Here, we apply a recency factor based on seven day volume in order to generate recommendations that are not just personalized to the wallet but also are relevant in the current period.
Our complete process of generating recommendations can be visually represented as a sequential pipeline of multiple stages as:
![Recommendation build flow](//images.ctfassets.net/lhkcghnr8ql4/5EGiB0BcHIcPmROPmiEysA/cf0f039bab9c87dc46e4a878190ffce6/image2.png)

## Conclusion
In conclusion, a reliable recommendation engine is important for the growth of the NFT industry and the marketplaces that operate within it. As more content producers enter the NFT space, it is becoming increasingly challenging for users to identify relevant collections that match their interests. This makes it hard for marketplaces to retain the attention and transactions of their users. 

Finally, at Upshot, we’re constantly iterating. This recommendation system will continue to improve with time. For example, we plan to add auxiliary information about wallets and collections, such as collection category, number of assets in the collection, asset traits, trait rarity, etc., to greatly enhance the recommendations. Furthermore, from the wallet’s perspective, we’re looking into ways to handle the  “cold-start problem” where we can still make accurate recommendations to a wallet that has not owned anything yet. 

As the NFT space continues to evolve, it is essential to have a tool that can keep up with the industry's changing landscape. Upshot's collection recommendations are an excellent step towards achieving this goal.


A reliable recommendation engine is important for the growth of the NFT industry and the marketplaces that operate within it. So, we built one. 

Introducing Collection Recommendations

### Launches and updates 🚀 🚀

![upshot-supports-more-collections](//images.ctfassets.net/lhkcghnr8ql4/4wWPcn9ooaNk4bi4oCgOmL/c8001a127e1717f3b50fd7b72f3660c8/upshot-supports-more-collections.png)

- **Launch:** With the appraisal scale-up efforts, we are now appraising over 67,000 collections with 74M+ assets at a sub-hourly cadence.
- **Feature:** Collection filtering on Orders tabs
- **Feature:** daily time series data released in endpoint

## Fixes and improvements 🔨

- **Improvement:** Optimized Infinite Scroll grids + NFT Card loading
- **Improvement:** Sidebar layout compatibility fixes + Table refactor
- **Improvements:** more alerts added (+ playbook started)
- **Improvements:** Support + downstream refactor work to handle appraisals scale-up / schema change
- **Improvements:** Airflow migration of ETH/USD away from ECS

Upshot Changelog [23-02-24]

### Launches and updates 🚀 🚀

![upshot-app-interface](//images.ctfassets.net/lhkcghnr8ql4/44P8KPwaMT2HYoPl2nn98L/3f40b3573c646c9c9aa7f51480cd36ab/upshot-app-interface.png)

- **Feature:** New sidebar component & associated layout changes
- **Feature:** New aggregator homepage landing
- **Feature:** Infinite scrolling wallet portfolio grids
- **Feature:** `/historical` time-series endpoint reconfigured + revealed

## Fixes and improvements 🔨

- **Improvements:** Revamped api tx logging, tiered-rate limiter, and related interceptor logic
- **Improvements:** Added endpoint ACL by tier
- **Improvements:** revamp/simplification of our existing alerts endpoints in preparation for public release. old endpoints marked as deprecated
- **Improvements:** alerts optimization (removing excess settings entity n topic)
- **Improvements:** add monitoring of alerts consumers
- **Improvements:** add timestamp filters to downstream part of alerts (ensures no events occurring >24hr ago are emitted as alerts)
- **Improvements:** Historical Appraisals (Daily Grain) for 120d
- **Improvements:** Pipeline migration for WEI -> ETH appraisal units
- **Fix:** Fixed blocking infra issue for the appraisal scale-up on staging with Data