When filtering lenders for a private equity deal, start with the criteria a lender cannot flex on: ticket size, sector appetite, EBITDA floor, geography, and product. Then layer in behavioral criteria: speed, flexibility on documentation, hold-vs-distribute posture, and track record with the sponsor. The goal is a shortlist of 8 to 15 lenders that can actually close the deal, not a long-list of names that might.
Most deal teams still build their lender list the same way. The lead partner names ten lenders, an associate adds five more, counsel adds three, and the list goes out. That process works until it does not.
Three shifts make this approach more expensive than it used to be:
The number of active private credit funds, direct lending platforms, and specialty finance groups has grown multiple times over the last decade. Naming the same 15 shops every deal means leaving real options on the table.
A shop that wins in software at say, $50mn to $150mn EBITDA is generally not the same one that wins in industrials at $20mn to $40mn. Treating lenders as fungible costs sponsors on pricing and flex.
Sponsors who track which lenders actually show up with strong terms, close on time, and behave well post-close build an advantage deal over deal. Sponsors who do not are starting fresh each time.
For more on how to calibrate list size, see this piece on finding the sweet spot for lender count on your deal.
The blocker is not intent. Sponsors know filtering would help. The blocker is that the information needed to filter well is scattered, unstructured, or missing entirely.
|
Source |
What is there |
Why it fails at filter time |
|
Lender tear sheets and pitch decks |
Stated check size, sector focus, and fund size |
Self-reported, often aspirational, rarely updated |
|
Past deal files |
Who actually funded, at what level, on what terms |
Locked in closed-deal folders, not queryable by criteria |
|
Banker and counsel intel |
Color on how a lender behaved on recent deals |
Informal, inconsistent, and not captured anywhere |
|
League tables |
Deal volume and ranking by product |
Backward looking, biased toward broadly syndicated, thin for private credit |
|
Deal team memory |
Context, relationships, lender personalities |
Walks out the door with the associate who had it |
The result: sponsors generally build lender lists from the pieces they can find quickly, not the pieces that would produce the best outcome.
A strong lender filter is not a single attribute. It is a layered set. Each filter narrows the universe to lenders that can realistically win the deal. Use these in sequence.
Every lender has a minimum and maximum hold size. A fund with a $200mn cap on a single position cannot anchor a $500mn unitranche. A specialty shop with a $50mn floor is not useful on a $25mn add-on financing. Filter first on check size fit, because this is the one criterion a lender generally cannot work around.
Some lenders run true sector teams (software, healthcare services, business services, industrials). Others are generalists. Sector-expert lenders generally move faster, ask sharper diligence questions, and price closer to the tight end of the market.
A lender that has already underwritten three SaaS ARR-based deals this year will get comfortable on a fourth in days. A generalist will usually take weeks.
EBITDA floors and ceilings matter independently of ticket size. A lender may have available capacity for a $150mn hold, but only in deals with $40mn plus of EBITDA. Below that, they pass.
Filter by the lender’s stated EBITDA band and by the actual EBITDA of their last ten closed deals. The second one is generally more accurate than the first.
Geography covers two things: where the borrower operates, and where the lender can lend.
Check jurisdiction fit before anything else. A lender that cannot lend into the borrower’s structure is a zero regardless of every other filter.
Match the lender to the product you actually need. Term loan B lenders, first lien direct lenders, mezzanine, and broadly syndicated arrangers all look like “lenders” in the rolodex and behave nothing like each other at the deal table.
Not every lender underwrites the same risk profile. Some shops anchor investment-grade corporates only. Others lean into stressed and distressed situations. Most sit in between.
Filter by the lender’s willingness to underwrite:
Lenders self-sort into leads, club partners, and participants. Some want to lead every deal. Others only want to join a well-formed club. A few will only participate quietly behind a lead.
Matching lenders to the right role avoids wasted conversations. A pure participant is not useful if you need someone to anchor structure and drive the process.
Prior relationship is one of the most underweighted filters. A lender that has funded three deals with the same sponsor generally moves faster, prices tighter, and flexes less on documentation. Relationship-based allocation is a real source of edge.
Filter by:
Some lenders can issue a commitment letter in ten days. Some take thirty. In a competitive sale, the difference is generally the deal itself.
Capture and filter by:
The filter does not end at close. Lenders behave differently once the deal is funded.
Pay attention to how each lender handles amendments, consent requests, covenant resets, and credit agreement flexibility once the company is operating. A lender that is easy at close and difficult during ownership is not a real long-term partner.
Apply the filters in this order. Each step should narrow the universe meaningfully.
The value of Termgrid for lender selection is not a single feature. It is the integrated capital markets workflow that captures structured lender data as a by-product of running deals on the platform.
Termgrid sits across the full CapMkts lifecycle. Data room, NDA, term sheet, lender communications, allocations and fees, capital structure data, and portfolio management all live in one place. Each closed deal contributes to a structured view of how each lender actually behaved, on what kind of company, at what hold size, on what terms.
That continuity is what makes filtering repeatable. The lender profiles hub, relationship insights, and lender engagement tools all draw from the same underlying record. With 30,000+ active users across 1,600+ lender institutions, most counterparties are already on the platform, so the data reflects real activity rather than self-reported tear sheets.
Use case. A mid-market sponsor is launching a $220mn unitranche for a healthcare services platform with $35mn of EBITDA.
Problem. The partner wants to run a tight process with 10 to 12 lenders. The deal team has a rolodex of 90 names, no easy way to filter, and a week before the teaser goes out.
Solution. The team applies hard filters (US-only, unitranche capability, $50mn to $300mn hold, $20mn plus EBITDA) and cuts the universe to 32. A sector filter (healthcare services track record) cuts to 21. A relationship filter (prior closed deals with the sponsor) surfaces 6 priority names. Behavioral filters rank the remaining 15.
Outcome. A 12-lender shortlist built on data rather than instinct. Faster teaser launch, stronger competitive tension, better terms at commitment. And because the work happened on Termgrid, every new data point sharpens the next shortlist without anyone having to rebuild the database by hand.
|
Stage |
How filtering gets used |
|
Pre-LOI |
Broad fit check on product, size, and sector to pressure-test the financing thesis |
|
Post-LOI, pre-launch |
Full filter pass to build the 8 to 15 name shortlist |
|
Teaser and commitment |
Behavioral filters drive which lenders get priority access and early calls |
|
Final selection |
Combine bids against relationship, flex, and documentation track record |
|
Post-close |
Capture lender behavior to sharpen filters for the next deal |
This feedback loop is what separates sponsors who run a structured process from those who start from scratch each time. Live market signal, such as financing grids on active deals, makes the loop tighter.
For most middle-market deals, 8 to 15 is the right range. Fewer than 8 reduces competitive tension. More than 15 generally signals a weak process and burns relationships on the lenders who do not win.
Product and check size fit, together. Everything else matters only among lenders who can actually fund the deal at all.
No. A fixed list ignores sector specialization, EBITDA fit, and current fund capacity. Each deal should produce its own shortlist.
Narrow to lenders with explicit distressed or special situations mandates, then filter on recent closed deals with comparable complexity. The list will be shorter, and should be. Track record on leveraged loans with stressed capital structures is generally a better signal than stated appetite.
Yes. Private credit filtering leans heavily on relationship, hold size, and sector fit. Broadly syndicated filtering leans more on arranger capability, distribution reach, and ratings appetite. Same framework, different weights on each filter.
Continuously. Every closed deal, every pass, and every post-close behavior is data. The quality of the next shortlist depends on the discipline of capturing the last one.
Termgrid
