S7 Grants Council Impact Analysis

[GFXlabs]

@thbialek

We applied a standardized attribution adjustment to each project’s observed TVL growth. This adjustment prorated impact based on (1) the relative size of other incentive programs running concurrently with the GC grants, and (2) the proportion of targeted pools or vaults relative to the protocol’s total TVL on the targeted chain. The attribution percentages applied ranged from 1% to 50%.

Can you explain the formula for this attribution share? They don’t make a lot of sense on the surface.

For example, Spark would not be deployed if not for the Grants Council grant. Yet 50% of TVL is not attributed.

The same logic was applied to Fluid on Superchain, another project where the grant was to get the deployment. (In this case TVL is currently 0, so it doesn’t change the numbers, but illustrates the point)

But let’s set the Attributed variable aside for the moment and address some major problems with these calculations.

It’s clear upon closer inspection that the reported numbers are, in fact, completely unusable. The TVL change does not actually measure the correct beginning date (which will be different for each project). The OP spent is overstated.

Let’s go through an example where this is obviously the case.

Bedrock was approved for a grant of 235k OP. Bedrock didn’t even begin to draft their application until March 26, clear intake until April 21, and be approved on until May 5. Zero OP was delivered until May 29. Yet we see here that TVL is being measured from March 20. The TVL changes need to be recalculated based upon some meaningful beginning date. Suggestions are OP delivery or announcement of campaign by the grantee.

Next, the OP attributed to this grant is being calculated on the entire grant rather than the OP actually delivered to the grantee. The denominator in dollars per OP needs to be recalculated.

Let’s use the provided data, even though the Attributed variable of 15% could be debated, and the increase in TVL is clearly measuring the wrong period. Just to illustrate how impactful this distinction is.

($8,520,000 TVL Increase * 0.15 Attributed) ÷ 235,000 OP Total Grant = $5.43 per OP

That’s very different from:

($8,520,000 TVL Increase * 0.15 Attributed) ÷ 94,000 OP Actually Delivered = $13.59 per OP

This is very important, because undelivered OP has not yet been spent. In fact, many grantees may not hit their milestones and never get the second tranche of funding at all.

As you can see, all of these numbers need to be recalculated using the correct beginning dates and the amount of OP actually delivered.

CC: @lavande who can help whomever did these numbers find relevant information, and to see if the SuperStacks, Futarchy, and other numbers likewise need recalculation