BIP-13: Adjustment to Weather Changes

Proposed: March 4, 2022

Status: Passed

Link: GitHub PR


Beanstalk Farms



The current Weather Changes are suboptimal for the current state of Beanstalk. In general, the Weather Changes are set in a conservative nature: Beanstalk would prefer to regularly pay a higher than necessary interest rate to attract creditors in a timely manner than regularly offer too low of an interest rate, fail to attract creditors, and enter a negative feedback loop. When the debt level is greater than Beanstalk is particularly conservative in its Weather Changes. BIP-9 radically improved the overall efficiency of the model but introduced new complexity around measuring demand for Soil, which has resulted in suboptimal Weather Changes. Now, as the price has once again started to oscillate more closely around 1 even with an excessively high debt level, it is an appropriate time for Beanstalk to be slightly less aggressive in raising the Weather when there is decreasing demand for Soil with 1 < P.

Proposed Solution

We propose the following adjustment to the Beanstalk Weather Changes in the cases when 1 < P and :

Technical Rationale

Beanstalk is implemented to support simple alterations to the Weather Changes. There are no technical modifications necessary related to the Weather Changes.

Economic Rationale

While the changes to the amount of Soil implemented in BIP-9 have been overall positive (e.g., the rate at which the Pod Rate was increasing has decreased dramatically), it also created an inefficiency in the measurement of demand for Soil, which in turn has resulted in some sub-optimal adjustments to the Weather. In particular, because the formula for the Soil at the start of each Season was changed to , most times there is a liquidity and time weighted average shortage of Beans in the BEAN:ETH Uniswap V2 over the previous Season, there is very little Soil available the next Season. Accordingly, when gas fees are considered, it is unlikely for anyone to sow Beans in the scarce amount of Soil, which makes measuring demand for Soil during that Season particularly difficult. Changing the 1 to a 0 in the case where there is decreasing demand for Soil with 1 < P will offer a rough temporary solution to this inefficiency.

To date, Beanstalk Farms has avoided proposing any short-term fixes to the protocol in favor of one-time upgrades that implement the desired final state of a given portion of the model. However, the cost to Beanstalk of refraining from making this change, (i.e., the Weather continuing to increase during most Seasons after 1 < P) while a better mechanism for measurement of demand for Soil is designed is likely to be significant over extended periods of time. Because the design for measurement of Soil is far from completed, it makes sense to make this short-term change.


Effective immediately upon commit.