Juan
Juan Defago

Protofire

The Graph Academy’s indexing expert Juan Manuel Rodriguez Defago has created this guide on comparing Indexers by evaluating their effective cuts. You can find him and his Graph-related repositories on GitHub.

Comparing Indexers

Assessing Effective Cuts to Compare Indexers.

The raw protocol cuts that Indexers display on The Graph Network cannot be used to compare Indexers directly with each other. In practice, this could mean that there are two indexers offering both a reward and query fee cut of 80% but their Delegators earn entirely different delegation rewards.

Let’s have a look how you can use effective cut percentages to better compare and evaluate Indexers in the following guide!

Understanding Effective Cuts.

The only cuts that can be compared equally are the effective cuts, since they are basically “normalized” for Delegations, which means that they are basically the cuts applied to the delegation pool.

If an Indexer has a negative cut, this means that the Indexer is giving away part of their rewards to match that number. In short, they are subsidizing their Delegators. For instance, this is done to attract new Delegators but has also been done by Indexers to make up for mistakes that they made in the past.

This is how a negative cut influences an Indexer’s profit margin:

Effective Cuts Explained with a Practical Example.

If the indexer has 20 GRT staked, and 80 GRT delegated, his total stake would be 100 GRT, and let’s assume it’s fully allocated, so the full 100 GRT are currently generating rewards.

Indexer’s Stake
20 GRT

Delegation
80 GRT

Total Stake
100 GRT

Let’s also assume that every 1 GRT staked generates 1 GRT in our math (in practice the numbers would be much smaller, but the ratios still match up). With all these assumptions, let’s check the case where our fellow indexer has a 10% reward cut, and how that translates to effective reward cut.

Our total daily rewards generated is 100 GRT (as per assumptions), with 20 GRT generated by the indexers own stake, and 80 GRT generated by the delegated stake.

Total Daily Rewards
100 GRT

Indexer Generated
20 GRT

Delegator Generated
80 GRT

Our indexer has a 10% reward cut, which means that he gets to keep 10% of the total amount of GRT generated. Since the total rewards were 100 GRT, he keeps 10 GRT, and gives 90 GRT to the delegators (Delegation pool). The delegation pool then distributes the rewards proportionally to every delegator in the pool based on it’s amount of shares of the pool.

Total Rewards
100 GRT

Indexer Reward Cut
10%

Indexer Rewards
10 GRT

Total Rewards
100 GRT

Indexer Reward
10 GRT

Delegator Rewards
90 GRT

So, our indexer generated 20 GRT, and kept 10 GRT, only 50% of what his own stake generated (This is what I call main stake reward generation in my spreadsheet FYI).

Our fellow delegators generated 80 GRT, and received 90 GRT, so 112.5% of what the delegated stake actually generated as rewards (This is what I call delegation reward generation in my spreadsheet)

If we then wanted to know the actual effective cut that the indexer applied, we can proceed to calculate based on it’s description: If the effective cut represents how much the indexer is taking from what the delegated staked generated as rewards, and our delegated stake generated 80 GRT (100%), but received 90 GRT (112.5%), then the effective cut is the difference between the percentages -> 100% – 112.5% -> -12.5%

Delegator Generated
100%

Indexer Generated
112.5%

Effective Cut
-12.5%

Example of a Positive Effective Cut.

Continuing on the example of effective cut, we could then check the positive effective cut scenario. Using the same scenario (20 GRT + 80 GRT):

Indexer Generated
20 GRT

Delegator Generated
80 GRT

Total Daily Rewards
100 GRT

But with 40% reward cut we can then see that:

Indexer Generated
100%

Delegator Generated
75%

Effective Cut
25%

Since both examples use the same ratios for own stake vs delegated stake (20 GRT – 80 GRT, 20%/80%), a higher reward cut will mean a higher effective cut. It’s important to note that if the ratio of own stake vs delegated stake changes, a higher reward cut might not mean a higher effective cut. That’s why reward cuts can’t be reliably used to directly compare one indexer to the next, since their delegation ratios might not be the same. Effective cuts on the other hand, can be directly compared, since they are directly related to the delegated stake, so it doesn’t care about any ratios.

Let’s change the first examples’ ratios.

Scenario: 10% reward cut, same as the original scenario, but indexer now has 5 GRT and the delegated stake is 95 GRT. The sum is still 100 GRT, and we assumed the same about the rewards.

Indexer Generated
5 GRT

Delegator Generated
95 GRT

Total Daily Rewards
100 GRT, 100%

With 10% reward cut we can then see that:

Indexer Generated
100%

Delegator Generated
94.73%

Effective Cut
5.27%

Summary

We’ve seen that with the same reward cut but a different ratio of owned vs. delegated stake, the effective cut is entirely different. In the original example, we had a -12.5% effective cut, the last one had a 5.27% effective cut. In the negative scenario delegators get a better APY, in this one they get a lower APY. Morale of the story: Reward cut is not directly comparable, effective cut is.

Go Back

Delegator Knowledge Hub