A pool can lose 40% of its TVL overnight with no warning, because nobody was tracking LP behavior at the wallet level. You assume the liquidity is stable. You have no data on who holds it, how long they have held it, or what conditions trigger them to withdraw. TVL looked fine until it did not.

If you are new to measuring DeFi health beyond TVL, the difference between TVL and active users is a useful starting point; the same tension applies to LP analytics.

Why TVL Alone Misleads DeFi Protocol Teams

Short answer: TVL aggregates capital without distinguishing the users and wallets behind it. Moving the unit of analysis from the pool to the individual wallets is what makes LP behavior data actionable.

Ten wallets holding $100M in a pool look identical in your dashboard to one whale holding $100M. The risk profile is completely different.

Protocols that rely on TVL as their primary health metric consistently run into the same problems. They cannot predict liquidity events because they have no visibility into LP concentration. They over-invest in incentive programs that attract mercenary LPs who exit the moment emissions drop, a dynamic covered in depth in Formo's guide to DeFi onchain retention. And they misread growth, celebrating TVL increases driven by a single wallet repositioning across pools.

LP behavior analytics solves this by asking: who holds the liquidity, how long have they held it, and what conditions cause them to remove it.

5 Key Metrics for Liquidity Provider Analytics 

1. LP Concentration Ratio

This measures what share of a pool's total liquidity is held by your top LPs. A healthy pool has distributed liquidity across many wallets. A fragile pool has 80% of its TVL concentrated in three wallets.

Track this as a rolling metric, not a snapshot. If your top-10 LP concentration is increasing over time, you are accumulating fragility even if TVL looks stable. Uniswap's v3 liquidity research found that the protocol attracted around 130,000 unique LPs, with two-thirds of positions held for more than one day. That distribution suggests skew: a relatively small share of active participants likely manages a disproportionate share of in-range liquidity, making concentration ratio a protocol-level risk signal worth monitoring weekly.

2. LP Retention Rate

LP retention measures the percentage of liquidity providers who are still active in a pool after a defined time window. Common intervals are 7-day, 30-day, and 90-day retention, the same framework used to measure DeFi user retention more broadly.

A pool with 60% 30-day LP retention has a fundamentally different business than one at 20%. The first has genuine product-market fit. The second is surviving on incentives.

The scale of the problem is significant. The DL News State of DeFi 2025 report estimates that cumulative DEX trading volume reached $11.4 trillion by end of 2025, up from $4.2 trillion at the start of 2024. That volume runs through LP-provided liquidity: every dollar of fee revenue generated at that scale flows from pools where LPs stayed. Even a modest improvement in LP retention translates directly to deeper liquidity and higher fee capture.

Calculate this at the wallet cohort level: group LPs by the week they first deposited, then track what share of each cohort is still providing liquidity at each subsequent interval. This is the same cohort methodology used in onchain user segmentation, applied to LP positions instead of general user activity.

3. Average LP Holding Duration

How long does a wallet typically stay in a position before withdrawing? Short average holding durations indicate mercenary behavior. Long durations indicate alignment with the protocol.

Segment this by wallet size. Smaller retail LPs are often stickier than larger wallets, which are frequently yield optimizers running automated rebalancing strategies. This segmentation changes how you design incentive programs.

4. Deposit and Withdrawal Velocity

Track the rate of new LP deposits versus withdrawals on a daily and weekly basis, normalized to exclude outlier whale movements. A sustained withdrawal trend at the wallet count level, even when TVL appears flat due to price appreciation, is an early warning signal.

Pay attention to withdrawal clustering: multiple wallets exiting within a short time window often indicates a coordinated response to an external event: a competitor pool launching, an incentive program ending, or a perceived protocol risk.

5. LP Wallet Quality

Not all LPs are equal. An LP wallet with a 90-day average holding duration, multi-protocol history, and no sybil flags represents a fundamentally different signal than a freshly-funded wallet chasing emissions. This is the core premise behind DeFi wallet intelligence, enriching raw onchain data with context.

Wallet intelligence adds DeFi experience level, cross-protocol activity, historical liquidity behavior, and wallet age to your LP data. This lets you segment LPs into archetypes rather than treating all capital as homogeneous.

How to Track LP Behavior Onchain

Short answer: Capture deposit and withdrawal contract events, aggregate at the wallet level, then enrich with activity history. These three steps cover the full LP tracking stack.

Start with smart contract event tracking

Every deposit and withdrawal to an AMM pool emits contract events. Setting up a contract event pipeline lets you capture LP activity at the transaction level: wallet address, amount, pool, timestamp, and fee tier. This is your raw data layer; from here you can derive all the metrics above.

Build wallet-level LP profiles

Aggregate contract event data at the wallet level to build LP profiles. For each wallet that has ever interacted with your pool, you want to know:

• First deposit date and most recent activity date

• Total capital deployed across all positions in your protocol

• Number of deposit and withdrawal events

• Average position duration

One important nuance: many LP wallets hold positions across multiple pools within the same protocol simultaneously. When a wallet reduces its position in one pool, it may simply be rebalancing to another pool rather than exiting the protocol entirely. Before flagging a withdrawal as churn, check whether the wallet is still active in other pools. Cross-pool LP view is the difference between measuring pool health and measuring protocol health.

Tools used at this layer include The Graph for event indexing and Flipside Crypto for analytics SQL. The wallet enrichment gap applies equally to all of them: raw event data shows what happened but not who did it and why. Formo's wallet profiles pull onchain history, in-app activity and behavioral data into wallet profiles your growth team can query and segment directly, without maintaining a custom pipeline.

How to Segment LPs with User Behavior Profiles

Once you have wallet-level data, segment LPs into cohorts based on behavior rather than capital size. The LP Behavior Profiles below gives each segment a defined profile, making it easier to benchmark over time and communicate health internally.

These segments tell different stories. Your loyal LP retention rate measures product health. Your mercenary LP share measures incentive dependency. Your new LP acquisition rate measures how well growth programs are working. For the full framework on acting on these segments, see Formo's guide to onchain user segmentation.

How to Use LP Analytics to Make Better Product Decisions

Incentive program design

Most DeFi incentive programs are designed without LP cohort data and end up subsidizing mercenary capital. If you track LP retention by cohort, you can measure whether a change in emission schedule improves 30-day LP retention or just temporarily inflates TVL. The DeFi growth experiments framework covers how to structure these tests so you get clean signal rather than confounded results.

Liquidity risk management

LP concentration ratios and withdrawal velocity give you early warning of liquidity risk. If your top-5 LP concentration climbs above a threshold while withdrawal velocity is also ticking up, that is a signal worth acting on before it becomes a liquidity crisis.

A useful rule of thumb: when your top-3 wallets collectively hold more than 50% of a pool's TVL, a single exit can realistically move aggregate TVL by 15 to 20%. Treat this as a calibration starting point; pool depth, asset volatility, and position structure all affect the actual impact. Set automated alerts on both metrics and review them weekly. Waiting for TVL dashboards to move is waiting for the crisis to have already started.

Partnership and BD targeting

Wallet intelligence lets you identify which LPs in your protocol are also providing liquidity to complementary protocols. These wallets are your highest-value BD targets for structured liquidity partnerships. This is a direct application of DeFi KPIs for growth teams: turning behavioral data into pipeline.

In practice, this means querying your LP wallet data for addresses that also appear in the contract interactions of protocols you want to partner with on Base or Arbitrum. Those wallets have already demonstrated willingness to provide liquidity in your sector. A structured incentive or co-liquidity arrangement targeted at this segment converts at a meaningfully higher rate than broad ecosystem outreach.

Common Mistakes in LP Analytics

• Measuring TVL at the pool level without breaking down wallet concentration

• Treating all LP capital as equivalent regardless of behavioral history

• Optimizing incentive programs based on TVL response rather than LP retention response

• Building Dune dashboards without wallet-level enrichment. You see what happened but not who caused it. (The Graph for event indexing and Flipside Crypto for analytics SQL have the same gap.)

• Looking at aggregate withdrawal data without clustering analysis, which misses coordinated exit signals

What Good LP Analytics Looks Like in Practice

A protocol team using LP behavior analytics effectively can answer questions like:

• What percentage of our LPs from the January cohort are still active today?

• What share of our TVL is held by wallets with a cross-protocol LP history of under 30 days?

• When we ended our emissions program in March, which wallet segments exited first?

• Which of our current LPs are also active in competitor pools on Base or Arbitrum?

Here is what that looks like in practice. A Base DEX notices its 30-day LP retention dropped from 58% to 34% between the January and March cohorts. Investigating the concentration ratio reveals that two wallets representing 44% of a major pool's TVL both exited in the first week of February.

The signal was visible earlier. Withdrawal velocity had been climbing for 10 days before TVL moved. With LP behavior tracking in place, the team sees this two weeks before the TVL dashboard fires an alert.

With that lead time, they can reach out to those wallets directly, adjust emission targeting toward the loyal LP cohort, and identify replacement capital from cross-protocol LP wallets already active in adjacent Base pools. Without it, they are reacting to a crisis instead of managing a risk.

These are not exotic questions. They are the questions that separate protocol teams making evidence-based liquidity decisions from teams that find out about problems when TVL alerts fire.

How to Get Started With LP Behavior Tracking

Short answer: Set up contract event tracking, aggregate at the wallet level, calculate retention cohorts, segment by the LP Behavior Profiles, then layer in wallet intelligence. The five steps below cover the full workflow.

If you are starting from zero, the practical path looks like this:

1. Set up contract event tracking for your pool's deposit and withdrawal events.

2. Aggregate events at the wallet level to build a basic LP activity table with deposit date, withdrawal date, and position duration per wallet.

3. Calculate LP retention by cohort for your first 90 days of data, grouping by deposit week and tracking survival at 7, 30, and 90 days.

4. Segment wallets by holding duration to identify your loyal versus mercenary LP mix using the LP Behavior Profiles above.

5. Layer in wallet intelligence to enrich LP segments with cross-protocol behavioral context: wallet age, DeFi experience level, and sybil risk signals.

Note: If your protocol is a lending market or yield vaults, your LP equivalent is the lender or depositor. The same methodology applies.

Formo is designed for exactly this workflow: contract event ingestion, wallet-level profile enrichment, cohort retention analysis, and user segmentation, without requiring your team to maintain a custom data pipeline. 

Spend less time on data plumbing, and more time making decisions with data you can trust. Learn more about how Formo approaches DeFi product analytics, or book a demo to see it applied to your DeFi protocol.

More in This Series

Exploring DeFi growth strategy? Read the other articles in this series:

FAQs

What is LP retention rate in DeFi?

LP retention rate measures the percentage of liquidity providers who remain active in a pool after a defined time window, typically 7, 30, or 90 days. Calculate it at the wallet cohort level: group LPs by the week they first deposited, then track what share of each cohort is still providing liquidity at each subsequent interval. A high LP retention rate indicates genuine product-market fit. A low rate indicates dependence on token incentives.

How do you measure liquidity provider concentration?

LP concentration is measured as the share of a pool's total TVL held by the top N wallets, typically the top 5 or top 10. A healthy pool has distributed liquidity. A fragile pool has 70 to 80% of its TVL concentrated in a handful of wallets. Track this as a rolling metric rather than a snapshot, since concentration can increase gradually even when TVL appears stable.

What is mercenary liquidity in DeFi?

Mercenary liquidity refers to capital provided by LPs whose primary motivation is short-term yield, typically token emissions or incentive rewards. Mercenary LPs are characterized by holding durations under 14 days and withdrawal patterns that correlate closely with emission schedules. This threshold applies most cleanly to DEX AMM pools. Lending protocol depositors and yield aggregator LPs often show naturally longer passive durations, so adjust the cutoff to match your protocol's typical engagement cadence.

What onchain data do you need to track LP behavior?

At minimum, you need deposit and withdrawal contract events from your AMM pool: wallet address, amount, pool identifier, timestamp, and fee tier. From these raw events you can derive holding duration, deposit and withdrawal velocity, and LP concentration. Enriching this data with cross-protocol wallet history adds behavioral context that raw event data alone cannot provide.