Decentraland Lords

Ryan Youngjoon Yi
Token Economy
Published in
8 min readMar 14, 2019

--

Source: market.decentraland.org

Real estate conversations in NYC nowadays will likely reference events such as the (non)-Amazon HQ2 in Long Island City, the (non)-shutdown of the L-train, and the continuous YIMBY-development high-rises in the neighborhood. The spillover effects of related events are well-documented.

Of course, environments in the real world and the value of its land are the products of compounding social and economic forces over history. So it is an interesting exercise to consider the real estate investment strategy of new developments — investing hefty prices to buy rights to greenfield land, attracting commerce and housing while developing the territory, and over time cultivate an identity and some community network utility effect.

Drawing that analogy to Decentraland, investors have bought rights to the $LAND and will enter the process of constructing utility into their assets with the soon-to-launch builder tools. I was keen to notice the patterns for digital real estate behaviors in a crypto-enabled marketplace.

Basics

  • As a refresher, each parcel of land in Decentraland is represented as an ERC-721 token ($LAND) with unique location parameters.
  • Parcels are meant to be interactive for their end users — the owners of the $LAND use tools like the SDK and builder package to create exploratory environments.
  • There are points-of-interest such as genesis plazas (green blocks), districts (blue blocks), and roads (light gray blocks). These are supposed to serve as common areas that are managed by different blocks in the community.
  • $MANA is the payment token in the Decentraland economy, though the functionality may change over time.

Timeline

Location & Block Parameters

By analyzing the smart contract metadata, we can see the relative positioning of $LAND assets that were auctioned off. The first auction sold a bulk of the central area (~34K $LAND), while the second auction skewed more towards the edges (~9K $LAND).

Auction 1 (Left), Auction 2 (Right)

[0] Auction Pricing of Location & Scarcity

[0A] — Auction 1

Given that world building is a key part of the Decentraland experience, I wanted to test out whether buyers took into consideration the “hot-spot” location proximity as a motivating factor. Let’s assume that landmarks (the roads, districts, and the genesis block) might serve as these points-of-interest. From the smart contract metadata, we can filter the price paid for parcels by their distance to the different classes of landmarks.

  • Filter parcels by distance to landmark type (road, genesis, district).
  • Map respective location distance as a ratio to the local maximum distance.
  • For instance, in Auction 1, the furthest parcel “distance” to the nearest road was ~60. A parcel with road location “1” is one block from the nearest road, and 1/60th of the maximum distance.
  • The higher the ratio, the further away from the landmark.
  • Calculate the median and average $USD contributed over the location parameter metadata, on log scale.

If proximity to landmarks was considered a driving factor, we would expect to see higher price paid for closer proximity. In other words, the higher the ratio, the less median or average $USD paid. We would expect a decreasing function to see a positive relationship between price paid for proximity. From Auction 1, we can eye-ball the linear trendlines for higher signal.

[0B] — Auction 2

If we were to run the same analysis for Auction 2, there is almost no relation between the location and price for all three landmarks. Yet, we see that the prices paid in Auction 2 versus Auction 1 are consistently higher for all location parameters across the landmarks. One reason could be the perceived scarcity of $LAND assets; another reason could be the project’s timing and execution risk priced into the assets (a full year of development and market activity).

[1] Marketplace Pricing of Location & Scarcity

While most of auctioned parcels (~90%) have not touched the marketplace, we can look at the marketplace to observe emergent behavior of the property assets values. Specifically, the buyer’s behavior might inform an assumption of his/her’s expectation of land value features.

We can cut this metadata in two ways:

  • Filter traded parcels by # of times traded on the marketplace (turnover).
  • Filter parcels by the price multiple relative to its auction price to infer perceived market value.

[1A] Turnover

One take-away from observing the turnover behavior is that $LAND parcels purchased on the secondary marketplace tended to be clustered. We also observe that parcels that were traded with high turnover often were within these cluster proximities.

Turnover Rate of $LAND Parcels

[1B]: Pricing Multiples

From the pricing filter, we can observe that the highest priced multiples (10x and above auction price) belonged to parcels that had relatively low turnover. This could imply that whales (or $LAND lords) perceived these parcels to be underpriced and followed a uniform pricing to inform their bids.

Pricing Multiples for $LAND Parcels

[1C]: Location

If we overlay the Decentraland atlas, we begin to see that the clusters form near the edges of landmarks. This might belie the purchaser’s expectation, from a UX perspective, that a user might begin by exploring the landmarks, and then venture into adjacent $LAND neighborhoods/parcels.

Turnover on Atlas
Pricing Multiple on Atlas

[2] Estate Pricing of Location & Scarcity

Estates refer to $LAND parcels that are adjacent to each other and aggregated as a single collection. The idea is that owning or building on an estate is like owning a small neighborhood — allowing the ability to create city-like experiences. Not surprisingly, the value of estates exhibit a positive linear relationship to its size.

$USD vs Estate Size

What about the marginal value of additional $LAND to an estate? Not so much…

  • Estate Price: Filter estates by size and calculate average price per parcel.
  • Base Price: Calculate the price of those individual parcels as the maximum of either the primary auction value or secondary market value.
  • Calculate the median of the premiums.

[2A]: $LAND Flipping

One early hypothesis before the analysis was that buyers would purchase adjacent $LAND parcels, aggregate them into estates, and then sell the estate. Interestingly, the data below shows that this behavior is true for market participants in the auction. We see that most estates sold were not comprised of parcels that were bought on the marketplace, implying that buyers bought these parcels $LAND originally in the primary, then flipped them as estates once the feature became available.

Estates vs $LAND Turnover
Estates vs $LAND Pricing Multiples

Although only a few estates are comprised of $LAND assets purchased on the secondary markets, Section [1B] may hint that buyers have formed estates that have not been sold in the secondary markets yet. It could be that these estate owners are waiting for the market to react to actual construction and building of these interactive environments before selling the $LAND estates.

[2B]: Estate Free Riding

In the estate diagrams, we see that there are many instances of individual parcels bought adjacent to estates. Although a more rigorous timeline analysis is required to dive deeper, it is likely valuable to own $LAND next to an estate, similar to the logic of sitting adjacent to a landmark.

Closing

The market behavior shown in this analysis assumes some expectation from the $LAND owners about the future usage and utility of their parcels (similar to many public token projects). The uniqueness of the Decentraland analysis is the attempt to understand NFT feature-specific variables that might drive different valuation characteristics from other fungible crypto assets. At a high-level these features boil down to proximity and size.

Decentraland is meant to be an interactive experience. With the builder tools imminent to launch, it will be interesting to track how $LAND owners will actually make their assets productive (building, leasing, buying adjacent, selling).We may see over time whether these land owners used their land productively or simply waited for favorable speculation. The findings could inform whether the distribution of $LAND via auction went to those who could most efficiently make use of the land (and thus bring utility and value to Decentraland), rather than rent-seekers, speculators or wealth-parkers.

As in most crypto use-cases, while enforceability of scarcity is strongly guaranteed, the parameters of scarcity are not perpetually fixed. In the case of $LAND, unlike real world estates, virtual world scarcity is not bounded by physics. Decentraland is an attempt to iterate on this idea. This may change over time via the usage of $MANA (largely ignored in this analysis) as a governance token for the Decentraland economy, which may eventually be used to vote on proposals related to scarcity and utility such as increasing the size of the map or the dimensions of parcels (re: the conundrum of governance…)

In virtual worlds we are collectively building an entertainment society in a vacuum minus the long-standing social and economic driving forces that create environments like NYC. Yet, we are all participating in the generation-long experimental phase of new incentives and environments, and so though Decentraland will take time to build as its own, we can at least track the utility in real-time. How long until my Decentraland estate gets gentrified?

All data was sourced via the $LAND smart contracts and nonfungible.com on 2019–02–28.

Disclaimer: The content provided on this site is for informational and discussion purposes only and should not be relied upon in connection with a particular investment decision or be construed as an offer, recommendation or solicitation regarding any investment. The author is not endorsing any company, project, or token discussed in this article. All information is presented here “as is,” without warranty of any kind, whether express or implied, and any forward-looking statements may turn out to be wrong. CoinFund and its affiliates may have long or short positions in the tokens or projects discussed in this article.

--

--