The lego stack / tower has been a captivating framework amongst those in DeFi. The basic notion is that several DeFi protocols can be combined in the form of a stack to provide different financial services. At the center of these combinations is the principle of composability, which dictates that protocols’ interoperable nature allows them to combine seamlessly. With the ease of protocol composability, product innovation can materialize via the combination of lego blocks.

Where does one see this today? Derivative protocols can build on top of Uniswap. In order to tap into deep liquidity for the underlying assets of the derivative instruments, these protocols may find it useful to plug into the well-known DeFi 1.0 AMMs. There are also fixed rate protocols working off of core money market functions delivered by Compound and Aave. A great example of this is Notional Finance, which makes use of Compound’s cTokens in order to build its yield flows. As these blocks combine, they create the potential for more combinations, inherently growing the stack. In some ways, going back to middle school math, adding one more object to the group leads to several new potential relationship subsets.

However, this piece does not go into the innovation catalyzed by combinatorics-based properties. Nor, and it is important to mention this, does this piece cover the risk characteristics exhibited by the lego stack. A pertinent question here revolves around how risks stack up / compound. My thoughts for this post revolve more around switching costs and their profile in the DeFi lego stack.

Switching costs have been a formative moat in financials and tech businesses. If one has ever gone over the core Morningstar moats, switching costs are right up there with brand value and economies of scale. What does one mean by switching costs? A high level interpretation is a set of financial, operational, economic, or other costs which arise when a business or consumer attempts to switch to a competitor. The switching cost could be considered negative to the end-consumer, but it is good for the business that is providing the service or product. Why? Because the stronger the switching costs, the harder it is for customers to leave… meaning more stickiness, retention, and opportunity to generate value.

Some great examples can be seen in Visa, Mastercard, and Salesforce. In the case of Visa and Mastercard, what drives the strong ROIC and compounding earnings growth of these two companies is the importance of their business to those that use it: banks and merchants. Becoming part of this network requires significant capital undertaking and commitment — banks can’t just change from Mastercard to a competitor in a couple of days. Visa and Mastercard run the piping infrastructure for financial transactions, and trying to remove one’s bank from that infrastructure is more than just rebranding, it is the alteration of a mission critical system. Thus, with switching costs, these two entities can count on strong retention and possess significant pricing power. In fact, both have consistently raised take rates over the past 10 years at a 3.5–5% CAGR.

The same can be argued for Salesforce. This company is a phenomenal SaaS business, especially as for the customer it is an easy entry (meaning it is not hard to start using and getting used to it), but a very hard exit. Moving off Salesforce after using it for a couple of years would constitute a major CRM headache — transferring data, profiles, logs, and more, in addition to changing key firm operations from sales to strategy to finance. Thus, the costs in terms of time / money as well as potential exposure to risks of data loss or business disruption make it hard for businesses to stop using Salesforce.

Overall, switching costs constitute a valuable moat which helps guarantee the reliability of revenue streams and customer retention, while also facilitating the opportunities for cross-selling and price increases.

Looping back to the intro, if composability provides the means for these lego blocks to seamlessly build off and combine with each other, then would that not lead to the flexibility to switch blocks? I think this can be the case in the short-term, however, in the long-term I am of the opinion that switching costs will strengthen, and consolidation will increase among the lower levels of the stack.

Two questions emerge from that opinion. Firstly, why, in the short-term, is flexibility possible and switching costs not too significant? Well, right now DeFi is still quite nascent. There have been several lego blocks created and introduced, and given that it is still early, the costs of switching are not too high. For example, a derivatives exchange that plugs into Uniswap could move to Balancer. Would it take some developer resources, audits, and testing? Yes, but it is feasible.

Why, then, as time goes by do switching costs strengthen? This is the more fascinating question to think about. A couple of points support switching costs strengthening:

1. Time. The nature of time strengthens a relationship as long as the blocks on top are happy with the lower blocks. In a way related to the Lindy effect, the longer lego towers stand on the blocks of Uniswap and Aave, per se, the harder it would be to change those blocks out. Hundreds of millions of dollars of liquidity would be in there, more code and technical debt would be tied to it, and psychologically, for the core contributors it just becomes something they are used to. If there is no terrible reason to switch, why undertake the effort and potential risks (especially smart contract-related) of moving over?

2. Lego-stack network effects. What happens when several of the mid-level blocks all interact on the same lower level blocks? Composability and functionality are at their prime. These protocols can work together in such an efficient and synergistic way that more value is generated being on those lower blocks than moving to others. In the case of Notional — what if a whole suite of financial services used cTokens? The ease of building a sub-ecosystem of products that allows for seamlessly moving from fixed rate lending / borrowing to other financial products is appealing to both the developers and users. As these lego blocks build and cross-combine, the lower blocks become more important as enablers of product combination.

3. Liquidity & virtuous cycles. If one pairs points 1 and 2 together, one gets this evolution whereby a couple of key protocols at the bottom comprise the foundations of most lego stacks. As more stacks are built on core DeFi 1.0 protocols, these have increasingly more liquidity. As these have more liquidity, it makes sense from an efficiency standpoint to build on them, hence increasing liquidity. As liquidity, and thus revenues, increase, so do: protocol resources, developer funding, community contribution, and product iteration. Thus, the protocol gets better (theoretically, of course). All of this makes it harder to switch. If one is managing a derivatives protocol that is plugged into Uniswap, which has the best liquidity and a solid developer team that is constantly making the product better, it becomes much harder to propose a switch to {other AMM}. The rationale for moving deteriorates. Simultaneously, the lower lego blocks consolidate into a few key protocols via the quest for efficiency / liquidity.

Thus, over time, switching costs should emerge in the DeFi stack. Mainly, they will be comprised of: decreased efficiency and liquidity, increased smart contract risk, and operational integration costs pertaining to sub-ecosystems.

These high level thoughts point to a pertinent conclusion: pay attention to the development of those lower level blocks. What is being built on top of them? Why are protocols integrating with them? How are integrations tracking? How are liquidity / AUM / listings changing? These blocks will have immense importance to the stacks of tomorrow. Understanding their progression towards building switching costs can be important to determining their viability as a crucial protocol of the future.

Disclosure:* This blog series is strictly personal/ educational and is not investment advice nor a solicitation to buy or sell any assets. It does not represent any views from where the author is working — all views, opinions, and arguments are the author’s. Please always do your own research.*