The statutory rate is the ceiling, not the floor, and most fund decks never explain what separates the two.

The bid compression in NJ specifically has been brutal for at least three cycles now, and anyone still modeling to statutory in Mercer or Essex county is just writing fiction. Your point about deed acquisition being "priced as an anomaly" is the one I want to pull on, because I have watched two fund operators get sideways on exactly that. Their back-office was built for a redemption-heavy world and then a 2020 to 2022 vintage portfolio started throwing off non-redeemers in counties where the deed process runs 18 to 24 months minimum and requires local counsel who actually knows the judge.

The fee stack problem you laid out is real but the duration extension is what kills the IRR math. A certificate sitting unredeemed for 36 months at 0.25 is not just low-yielding, it is tying up capital that the model assumed would recycle into new certificates. The fund's effective deployment rate collapses and the GP is still collecting the AMF on committed or deployed capital, depending on how the LPA reads, while the LP watches recycling assumptions fall apart in real time.

I am on the capital side, not running a lien fund, but I have looked at three decks in the last year where the county-level clearing rate data was either absent or buried in an appendix with a footnote. What county mix are you actually underwriting against, and are you getting auction results from the county clerks directly or through a data vendor, because I have seen vendor data that lags badly enough to matter?

We modeled to statutory on our first small raise and got burned by exactly this. Actual clearing rates in the counties we targeted came in at 0.25 to 1 percent, and the fee stack ate the return before we could even talk about distributions.

The deed acquisition piece is where I have never seen a fund actually perform the way the deck describes. They always model it as upside, a bonus scenario, but when you are understaffed and a certificate goes to lien foreclosure in a county that requires strict compliance on notice and timing, that "upside" turns into 18 months of legal cost and a property nobody wanted.

Are you seeing any GPs in this space underwriting to clearing rate by county and then stress-testing fee coverage at those levels, or is everyone still backward-engineering a haircut from 18 and hoping redemptions cluster early?

The deed acquisition piece is the one that actually ends funds, not the yield compression. I watched a group out of Atlanta spend fourteen months and legal fees they never budgeted trying to clear title on three Georgia certificates they never expected to foreclose on, and it ate the carry on the whole vintage.

On the underwriting question: anyone still applying a haircut to statutory instead of just pulling the actual clearing data from the county is modeling a fiction and charging real fees against it. DeKalb auctions I have watched were clearing sub-one percent last year on anything with clean title, and that is the number you start from, not 18 minus a gentleman's discount.

What county mix does your target fund actually hold, because NJ is not one market?

What's your fund's target county list, and have you pulled the last four quarters of clearing rates for each one individually?

The funds that model to statutory and apply a haircut are still wrong, just slightly less wrong, because the haircut is usually a judgment call rather than a county-by-county clearing rate pull from the last two or three auction cycles. Underwriting to actual clearing rates by county and lien type is the only number that holds, and even that needs a redemption curve assumption baked in, because 0.25 percent on a two-year redemption and 0.25 percent on a six-year non-redemption that triggers deed work are not the same investment at all.

The fee stack problem you identified is the real structural issue. A 1.5 percent AMF against 3 to 4 percent net yield at competitive clearing rates is not a fee, it is a transfer from LP to GP that the statutory rate fiction makes look reasonable in the deck.

Are you seeing any funds that have actually restructured the promote to be redemption-speed-adjusted, so the GP earns more when certificates turn fast and less when they age into potential deed situations, or is the standard structure still just a flat promote against total deployed capital?

The question about county-level clearing rates is the right one, but the more surgical version is whether the underwrite separates premium counties from distressed-municipality counties within the same state, because those two pools clear at rates that can differ by 10 percentage points and they get averaged into a single portfolio yield figure that flatters neither. A fund deploying across Bergen and Cumberland in the same vehicle is not running one strategy, it is running two with different redemption timelines, different deed-acquisition probabilities, and different servicer demands, and a blended statutory haircut obscures all of that.

The duration point in your post connects directly to fee drag in a way that compounds the problem. At 0.25 percent on a certificate that redeems in month 14, the gross return on that position is roughly 0.29 percent total. The 1.5 percent AUM fee accrues annually on deployed capital regardless, so the fee on that single position exceeds the interest earned by a factor of five before the fund has paid a dollar of admin. The math only works if non-redemption rates are high enough that deed acquisitions generate returns that cross-subsidize the performing certificates, which means the fund's actual thesis is real estate acquisition, not lien yield, and the deck should say that.

What I want to know is what your fund's actual non-redemption rate has been over the last two vintage years, because that number determines whether the deed-acquisition path is an emergency protocol or the implicit return engine.

Deed acquisition nearly broke a deal I was watching from the outside in Knox County, took fourteen months and two attorneys to clear title.

The fee stack surviving when the yield doesn't is the thing that should end conversations with GPs and mostly doesn't.

I got burned on the duration side on a single certificate in Cumberland County, redeemed at month 26 at 0.25, and the carry cost on my slice of the admin made it a wash at best. Not a fund, just a direct position, but same math at scale.

Are any of the funds you're looking at publishing realized yield by county after fees, or is it all IRR on the deck with the clearing rate assumptions buried in an appendix?

The number I would want before anything else is your projected redemption rate by county, because it is doing more work than the clearing rate itself. A certificate that bids at 0.25 and redeems in six months is a different animal than one that bids at 0.25 and sits for three years before the fund has to decide whether to pursue deed acquisition, and most fund models I have seen treat redemption timing as a single blended assumption rather than a distribution.

What is your fund's current staffing or servicer arrangement for deed acquisition, and has that been costed into the carry structure, or is it sitting off to the side as a deal-by-deal expense?

Underwriting to clearing rate requires county-level auction data for at least four consecutive cycles before you commit capital to that geography.

Underwriting to clearing rate only matters if your deed-acquisition pipeline is funded and staffed before the first certificate expires.

What does your fund's operating agreement say happens when redemption falls below a threshold, and does it give you discretion to pursue the deed or require LP consent?

Wholesaled one tax lien adjacent deal in Memphis and the seller's attorney kept referencing the statutory rate like it was actual expected yield, and I had to pull the last three auction results myself to understand why the numbers didn't add up.

What nobody above mentioned is that the county mix inside a single fund can swing the clearing rate dramatically, so "we underwrite to auction" is almost meaningless without knowing whether those auctions are Passaic versus Cape May versus something in a state with totally different competitive dynamics.

Are your target counties the ones where institutional money is already saturating bids, or are you finding pockets where retail bidders still set the clearing price?

The duration mismatch you're describing at the end is the part I'd push on harder. A 0.25 percent certificate that runs two years and redeems is annoying, but a fund that hits 15 percent non-redemption and wasn't staffed for title work is a capital event, not a performance footnote.

What's the fund's disclosed assumption on redemption rate, and is it county-specific or a portfolio-level blended guess?

Maricopa probate, so different animal, but my private lenders want a number not a range.

Redemption timing compounds the duration problem in a way most models miss: a certificate redeeming at month three at 0.25 percent is not just low-yield, it triggers redeployment friction, and if the fund is paying management fees on committed capital rather than invested capital, the LP absorbs the drag twice.

What county are you seeing the sharpest spread between statutory and clearing rate right now?

Two things in that post I want to push on. The duration piece is right but the framing undersells it. A 0.25 percent certificate that sits unredeemed for three years and then tips into a tax deed process in a county with a contested chain of title is not just an underperformance story, it is capital tied up that the waterfall never modeled. Most decks I have seen show a redemption distribution centered somewhere around fourteen to twenty-two months, which already feels optimistic for distressed owners who could not pay the original bill.

On the underwriting question: are you seeing funds that will actually share county-level clearing rate data from the prior two or three auction cycles, or is it all blended portfolio projections that let them hide New Jersey inside better-performing states?

Modeling to clearing rate is the right frame, but the number that actually matters inside it is the expected redemption timeline distribution, not just the rate. Take a 10 million portfolio at a 0.25 percent clearing rate with an average redemption at 18 months. The gross yield on the lien position is roughly 25 basis points annualized, so around 37,500 dollars on the whole book before fees, servicing, or carry costs. The 1.5 percent AMF alone is 150,000 dollars. That math does not need commentary. What makes it worse is that the timeline distribution is skewed: fast redemptions at 0.25 percent contribute almost nothing, and the tail positions that run long enough to matter on a time-weighted basis are exactly the ones heading toward deed acquisition, which carries its own legal and remediation cost stack that most offering documents treat as a footnote.

The piece I would push on is how the fund is underwriting the non-redemption population specifically, because that is where the return variance actually lives at compressed bid levels. If a certificate does not redeem, the fund's total return depends almost entirely on what the property is worth net of foreclosure costs and any municipal super-priority claims that survived the sale, and those figures are county-specific and parcel-specific in ways that a portfolio-level clearing rate haircut cannot capture. I would want to see the fund's historical redemption rate by county and vintage, their average deed acquisition cost per parcel, and what percentage of deed acquisitions resulted in a resale above cost. If they are modeling the non-redemption population as an upside event rather than a liability with conditional probability, that is the assumption doing the most work in the deck. What does their offering document say about deed acquisition reserves, and are those reserves funded at close or drawn from LP capital calls as inc

Underwriting to clearing rate is the only defensible starting point, and the funds still on statutory with a haircut are just putting the guess in a different place where it is harder to audit. The haircut number tends to drift toward whatever makes the IRR slide look acceptable, whereas a county-level clearing rate average, even a trailing twelve months from public auction results, is at least falsifiable.

The duration question you raised is where the two approaches actually diverge in practice. If you model to clearing rate, a two-year non-redemption is already a bad outcome baked into a low-yield certificate, so the deed pathway has to be underwritten as a real exit with real costs attached, not a footnote. Funds modeling to statutory tend to treat non-redemption as a windfall because the foreclosure upside looks large relative to the modeled yield, which means they are neither staffed nor capitalized for the cases that actually need title work, eviction, and a disposition strategy.

What county mix are you looking at, and are you seeing clearing rates stabilize anywhere or still compressing quarter over quarter?

Staffing for deed acquisition as an "anomaly" is the line that got me. That's not a modeling choice, that's a fund manager who has not worked a lien portfolio through a down cycle and does not know it yet.

What's the redemption assumption baked into the return waterfall?

Underwriting to clearing rate by county is the only way to do it, but even that breaks down if you're pulling comp data from prior quarters and bid compression is still moving.

What I keep tripping over in these decks is that deed acquisition gets footnoted as tail risk when it's actually the scenario that tests whether the GP has real operations or just a spreadsheet. Do you know what percentage of the fund's target certificates historically go to deed in those specific counties?

Underwriting to statutory with a haircut still puts you in the wrong place because the haircut is usually applied to income, when the real variable is bid level by county and certificate class. A 10 percent haircut on 18 gets you to 16.2, which has no relationship to a clearing rate that is already sitting at 0.25 on residential paper in competitive New Jersey counties. The model needs to start at observed clearing rates for each certificate type and work forward, not start at statutory and work backward.

The duration point you raised is the one most fund models treat as binary when it is actually a distribution. Confirm with your fund counsel how the LLC operating agreement defines the deed acquisition pathway and whether that process is capitalized separately, because many agreements are silent on the capital call mechanics that follow a non-redemption, and LPs who signed on for a lien fund do not always have appetite for an entity that is now managing distressed residential assets.

Are you seeing differentiation in clearing rates between residential and commercial certificates at the same auction, or is compression roughly uniform across certificate class?

We underwrote to county-level clearing rates on our last raise, and the number that broke two co-investors was not the bid rate itself but the variance across counties in the same state. Same statutory ceiling, 40-basis-point spread between what cleared in Essex versus what cleared in Cumberland, and both of those were moving quarter to quarter. A single haircut applied to 18 percent does not capture that, it just hides the modeling laziness inside one number.

What's your redemption rate assumption, and is it tiered by certificate age or flat across the portfolio?

The duration mismatch you are describing has a specific clause problem underneath it: most offering documents define "projected return" against the statutory rate without a corresponding representation about expected bid levels, which means the GP is not in breach when yields compress, they just did their underwriting poorly and the LP bears it. Take a 10 million fund modeling at 18 percent gross, with a 1.5 percent AUM fee and a 20 percent promote over an 8 percent preferred return. If the portfolio clears at 3 percent, the pref is not met, the promote does not trip, but the AUM fee still runs, and the LP is netting something closer to 1.5 percent on deployed capital before administrative drag. That is not a bad fund quarter, that is the structural outcome of a fee stack that was never stress-tested against competitive auction conditions.

The deed acquisition exposure compounds this because most operating agreements treat property acquisition as a discretionary act and budget for it as a percentage of the certificate pool, typically 2 to 5 percent going delinquent. In a compressed-yield environment where sponsors bid aggressively to deploy, they often concentrated in lower-value residential paper where redemption rates are actually lower, so the 2 to 5 percent assumption can easily double, and the fund is suddenly a distressed residential operator with no underwriting for that role and no fee structure that compensates the GP to do it well. The GP is incentivized to move slowly on deeds because the AUM fee keeps running on the certificate balance.

The question I would put to any sponsor before the wire: show me the weighted average clearing rate from your last three auction cycles by county, not the statutory ceiling, and show me your deed acquisition staffing and the timeline assumption built into the model for properties that do not redeem within 24 months. If those numbers

The point about duration risk and deed acquisition being correlated failures is the one worth pulling on further. A fund that underwrites to clearing rates is still exposed if it treats redemption timing as fixed, because the same distressed property that bids down to 0.25 percent at auction is often the one where the owner is least able to redeem on schedule, so the portfolio's lowest-yielding certificates are also its highest conversion candidates.

The staffing piece is actually a capital allocation question disguised as an operations question. Deed acquisition requires title work, quiet title in some jurisdictions, and often property management or disposition capacity, and those fixed costs have to be spread across however many conversions the portfolio produces, which at scale with compressed clearing rates could be a meaningful share.

What does your fund's LLC or LP agreement say about the expenses associated with deed acquisition, specifically whether those run through the fund or get charged to the GP's management entity?

The underwriting question cuts to the mechanics pretty directly: if you model to auction clearing rates, you need county-level bid data broken down by lien size, because small-balance liens in rural counties often clear at meaningfully higher rates than the institutional paper in the big suburban counties where most fund capital concentrates. A fund deploying at scale gets pushed toward the competitive tranches almost by gravity, so the effective portfolio yield is not even the average clearing rate across all counties, it is closer to the clearing rate in the tightest markets weighted by where the capital actually lands. A 1.5 percent AMF against a 3 percent gross yield is not a rounding error, it is a structural misalignment, and the promote hurdle set at 8 or 10 percent preferred looks reasonable on the cover page and becomes nearly unreachable once you account for servicing, legal on contested redemptions, and the carrying costs on any certificate that ages past the typical two-year window.

The duration piece you raised connects to something the fund documents rarely resolve cleanly, which is what the waterfall looks like when deed acquisition becomes a real outcome rather than a footnote. Most LP agreements I have seen treat the certificate and the eventual deed as the same asset class for distribution purposes, but they are not the same operational or capital event. The deed acquisition phase requires title work, municipality negotiations, and sometimes eviction or quiet title litigation, and the timeline from certificate to clear title can stretch three to five years in contested cases. If the fund has a fixed term, those assets either exit at a discount or the GP requests an extension, and most LP agreements give the GP that extension right with fairly light consent thresholds. The LP is then locked into an asset that is performing at a fraction of the modeled rate, in a structure that was sized for a different environment. The number I would want before anything else is the fund's

The duration mismatch compounds in a way fund decks almost never model explicitly: certificates that redeem fast do so because the borrower had liquidity or refinanced, which skews your early redemption pool toward the least distressed properties, and the liens that sit are the ones on assets with title problems, estate disputes or deferred maintenance that suppresses any eventual deed value. So the fund ends up holding a tail that is both slower and messier than the average certificate in the portfolio would suggest, and the capital allocated to that tail is idle or performing at 0.25 percent while the deed process grinds. If the fund modeled duration at 18 to 24 months to inform its capital recycling schedule, that tail is also breaking the redeployment thesis.

On your actual question: the funds I find credible at this scale underwrite to trailing 12-month auction clearing rates by county and then stress-test against a scenario where 20 to 30 percent of the book does not redeem inside 36 months, which forces them to budget for the legal and administrative cost of deed acquisition before it happens rather than treating it as an edge case. I would want to confirm the specific county-level clearing data directly from the municipal tax sale records, not the fund's own summary, and I would want to see the servicer agreement to understand whether deed pursuit is handled in-house, farmed to a specialist, or simply not contemplated in the ops budget. The gap between those two answers tells you whether the fund is managing a real portfolio or a modeled one. What does your servicer arrangement actually cover once a certificate hits the two-year mark without redemption?

Underwriting to clearing rate, full stop, no haircut fiction on top.

Modeling to actual clearing rates by county for the last six months, not statutory minus a haircut.

The assumption doing the most work in your post is that the fund modeled to statutory and applied a haircut, which treats the 18 percent as the anchor. The more dangerous version is a fund that modeled to statutory, got LP commitments, and then chased volume to deploy capital fast enough to satisfy the pacing expectations in the deck. When you are under pressure to deploy 10 million in 12 months across competitive New Jersey counties, you are not optimizing for yield at auction, you are optimizing for certificate count, and the bid behavior reflects that. The clearing rate at 0.25 percent is partly a product of fund capital that has to move.

The deed acquisition piece connects directly to that deployment pressure in a way the OM rarely captures. Most fund agreements define the manager's discretion over post-certificate actions in language that favors the manager on timing. The decision to pursue foreclosure, which is the mechanism that converts a non-redeeming certificate into a deed, typically sits with the GP, and the fee structure usually attaches a separate acquisition fee at that point, often 1 to 2 percent of the property value, on top of the asset management fee that was already running during the redemption period. So the fund underperforms on yield, and then the resolution of the underperformance generates a second fee event for the manager. The LP documents often permit this because nobody modeled a portfolio where non-redemption was common. What counties are you seeing the worst clearing rates in right now, because Essex and Hudson behave very differently from Cumberland in terms of property type and realistic deed-path economics?

The duration mismatch point is the one I keep staring at in these memos. Most of the ones I have read treat non-redemption as an edge case at maybe 5 or 8 percent of portfolio, but if you are buying in the competitive NJ counties where bid compression is worst, you are probably buying liens on properties where owners have the least liquidity to redeem, so those two risks are positively correlated in exactly the wrong direction. I have seen one memo in the last year that actually modeled to county-level clearing rates rather than statutory, and they were using Essex County 2023 results averaging somewhere around 0.5 to 1.2 percent depending on property class, which at least felt honest. The fee structure still made me nervous because 1.5 AUM on that yield range is not a rounding error, it is the return. What is the average certificate face value in the portfolio you are looking at, because that changes the deed acquisition staffing question a lot, a fund full of 8,000 dollar liens on vacant lots is a different operational nightmare than one concentrated in 200,000 dollar residential liens that are more likely to redeem anyway.