The Real Risks of Investing in Hard Money Loans, From a Lender Who Has Foreclosed

Every risk in private lending, named plainly by a lender who has taken properties back: default, value decline, illiquidity, fraud, leverage and manager risk.

Apurva Sanghavi · October 2, 2026 · 20 min read

Why a lender is writing this

Most articles about the risks of private lending are written by people who have never had to stand in a half-finished house and decide what it is worth. I have. Since October 2019 we have funded 60 Texas loans totalling $13.3 million, with 55 repaying on schedule and five going to workout: four foreclosures and one deed in lieu. Three of those five returned principal in full. Two did not. Total principal lost across all 60 loans has been under $50,000, which is less than 0.4% of principal funded. Those figures are the manager's own lending record and are not audited fund performance. I include them here because a risk article from a lender with no losses to discuss is a marketing document, and because what I learned in those five files is the only part of this post that cannot be found elsewhere.

Risk 1: the borrower stops paying

This is the risk everything else descends from. A borrower misses a payment, then another, and the loan that was an income stream becomes a project.

The early signals are consistent enough to be useful. A payment arrives late rather than short. A borrower who answered the phone in week one stops answering in week six. Requests for a payoff quote stop coming. A construction draw request arrives with invoices that do not match the work visible on site. An insurance policy lapses and the notice comes to us rather than from the borrower. A property tax bill goes unpaid, which in Texas matters more than in most states because the taxing authority's lien sits ahead of ours.

What a servicer does in week one is unglamorous and decisive. Call, email, send the written notice required by the loan documents, confirm the insurance is in force, confirm taxes are current, and get a straight answer about why. Most of the time the reason is ordinary: a sale fell through, a contractor walked off, a refinance was declined. Many of these resolve with a short extension and a fee, and a lender who panics in week one converts a solvable delay into a foreclosure.

By month three the tone has changed. The file moves from servicing to workout. We order a current valuation rather than relying on the one from origination. We inspect the property, which often means going ourselves. We check for new liens, unpaid taxes, mechanics' liens from unpaid subcontractors and any bankruptcy filing. We instruct counsel to begin the notice process while still negotiating, because the two run in parallel and the notice period cannot be shortened later.

The decision at month three is rarely between good and bad. It is between a modification that reduces the chance of full recovery and a foreclosure that takes months and costs money. There is no version of this where the loan simply pays.

Risk 2: foreclosure takes time and costs money

Texas is one of the faster states for a lender, and it is still slow.

Most investment-property loans here are secured by a deed of trust with a power of sale, which permits non-judicial foreclosure. Under the Texas Property Code, notice of the sale must be given and posted at least 21 days before the sale date, and sales take place on the first Tuesday of the month at the location designated by the county. Loans secured by a borrower's residence carry an additional notice of default and opportunity to cure before the sale notice can be given. This is a description of the general framework, not legal advice, and any individual file needs counsel.

So the theoretical minimum, from decision to sale, is about a month. The realistic timeline is longer, because the decision is rarely made cleanly, because borrowers negotiate, and because a bankruptcy filing on the Monday before a first-Tuesday sale stops everything. That last one is common enough to plan for. The automatic stay halts the sale, and the lender then spends months and legal fees seeking relief from the stay while the loan earns nothing.

The costs stack up while the clock runs. Attorney's fees for the notice process and any contested proceeding. Posting and trustee fees. Property taxes, which keep accruing and must usually be paid to protect the lien position. Force-placed insurance once the borrower's policy lapses, which is more expensive than ordinary coverage. Utilities, securing the property, and often a lock change and a clean-out. After the sale, there may be an eviction. Then brokerage and closing costs on the resale.

No trade body publishes a single all-in percentage-of-balance figure for this, because the number depends heavily on contested versus uncontested files and on how long a bankruptcy stay runs. What is documented: Texas Property Code §51.0075 caps a substitute trustee's fee at the lesser of 2.5% of the sale price or $5,000, and Fannie Mae's servicing guide (allowable foreclosure fees, Texas, 2024) caps the combined attorney, trustee and notary fee for an uncontested non-judicial foreclosure at $2,300. Add accruing property taxes, force-placed insurance, and resale brokerage and closing costs on top of that base, and a contested file with a bankruptcy stay can run well beyond it in both time and legal fees. Budget for the uncontested case and expect the contested one.

The condition a property comes back in is the part nobody warns you about. A borrower who has stopped paying has usually stopped maintaining. Materials disappear. Copper disappears. A half-completed renovation is worth less than the house was before the renovation started, because the next buyer has to undo the work as well as finish it. In one of our own files, a new construction single-family loan in Houston 77020 for $176,400, funded in January 2022, we foreclosed and recovered principal. The recovery came from the fact that the land and the partially built structure retained value in that submarket, not from anything clever we did after the default.

Risk 3: the collateral is worth less than you thought

A first lien is only as good as the number it sits behind. Three ways the number goes wrong.

The valuation was wrong at origination. An appraisal is an opinion produced by a person who was paid by someone with a preference about the answer. Comparable sales can be stale, adjusted aggressively, or drawn from a different micro-market a few streets away. In a rising market, bad valuations hide; in a flat one, they surface all at once.

The market moved. A loan written at 70% of value needs roughly a 30% decline before the principal is bare, and that is genuine protection. It is not unlimited protection, and it shrinks as accrued interest, unpaid taxes and foreclosure costs are added to the balance. A 70% loan with nine months of unpaid interest and a tax bill can be at 80% of a value that has itself fallen.

The renovation was never finished. This is the most common and least discussed, and it is the one I would put first. In a fix-and-flip loan, the underwriting usually relies on an after-repair value. That value assumes the work gets done. When a borrower runs out of money at 60% completion, the property is not worth 60% of the way between purchase price and after-repair value. It is often worth less than it was on the day we funded, because it is unsellable to a retail buyer, the systems are open, and the only buyers are other investors who price in their own profit. Underwriting to as-is value and releasing money against inspected draws is the defence, and it is why draw discipline matters more than the headline loan-to-value ratio.

Risk 4: your capital is locked

This risk has nothing to do with borrowers and is the one investors underestimate most.

Matador Note Fund 1 LLC has a 24-month lock-up. After that, redemption requires a written request with 90 days' notice, is processed quarterly, and is capped at up to 25% of your balance per quarter, subject to fund liquidity. Read that last clause carefully, because it is the one that binds. A fund whose loans are not paying off cannot create liquidity to meet redemptions, and nearly every private fund's documents contain a mechanism allowing the manager to slow or suspend redemptions when honouring them would harm remaining investors. That mechanism exists for a sound reason and it is still a gate.

Work through the timing. Request redemption in month 25, serve 90 days' notice, get processed at the next quarterly date, receive up to a quarter of your balance, then repeat. Returning a full position can take a year of quarters after the lock-up ends, in good conditions. In bad conditions it takes longer, and bad conditions are exactly when people want their money.

There is no secondary market of any consequence for these interests. Assume that what you commit is unavailable until the loans repay, and treat any earlier liquidity as a bonus rather than a plan.

Risk 5: leverage

If the fund borrows against its own loans, the arithmetic runs in both directions and not symmetrically. A facility raises returns on equity when the loans perform and the spread is positive. It deepens losses when they do not, for one structural reason: the facility lender is repaid before fund investors are. Borrowing-base tests and advance-rate cuts can also force a fund to pay down debt at the worst possible moment, sometimes by selling performing loans at a discount.

Matador Note Fund 1 LLC may use a note-on-note credit facility within the limits set by its Private Placement Memorandum. The full arithmetic, in both directions, is set out in our post on leverage.

Risk 6: fraud

Fraud in private lending is not exotic. It is a small number of patterns, repeated.

Straw borrowers. The person on the note is not the person controlling the deal, usually because the real principal has a history that would fail underwriting. Controls that catch it: identity verification, independent credit and background checks on every guarantor, questions about the source of the down payment, and asking who is actually managing the project on site.

Forged or diverted payoffs. A payoff statement is intercepted or altered and funds are wired to the wrong account. Controls: callback verification to a known number, never to one in the email; dual authorisation on outgoing wires; using established title companies and confirming their wire details independently.

Double-pledged collateral. The same property is pledged to two lenders, exploiting the gap between funding and recording. Controls: title insurance with a lender's policy, funding through a title company or attorney, a gap indemnity, and confirming recording rather than assuming it.

Contractor theft and draw fraud. Invoices for work never performed, lien waivers that are not real, materials paid for and never delivered. Controls: inspected draws with photographs, paying against verified completion rather than invoices, direct payment to suppliers where warranted, collecting conditional and unconditional lien waivers, and running a title update before later draws to catch mechanics' liens.

No control set eliminates fraud. A determined and competent fraudster with a cooperative insider defeats any of the above. The realistic goal is to make the common patterns fail cheaply and to catch the rest early enough that the collateral is still worth something.

Risk 7: the manager

You are not buying loans. You are buying a manager's judgement, applied repeatedly, in conditions you will not see.

Key-person dependence. In most small private funds, one or two people make every credit decision and run every workout. If that person is gone, the process that produced the record is gone. Ask what happens if the principal is incapacitated, who has signing authority, and whether there is a written succession arrangement.

Servicing quality. Collections, escrow, insurance monitoring, tax monitoring, inspections, draw administration. This is unexciting work that determines outcomes. A lender who does not know a policy lapsed in March finds out in July when the house burns.

Conflicts with the manager's other businesses. A manager who also originates loans outside the fund has to decide which entity gets a given loan. A manager who also invests in real estate may compete with borrowers. A manager taking an asset management fee on invested capital has an incentive to deploy. Our own arrangement is that the fund pays a 2% annual asset management fee on invested capital, the 8% preferred return is paid to investors before any manager profit share, and above that excess distributable cash is split 60% to investors and 40% to the manager. That structure aligns some interests and not others, and you should assume every fee structure in this industry does the same.

The single best test of a manager is how they talk about their losses. Ask for the worst loan they ever wrote. If the answer is short, specific and unflattering, keep listening. If there has never been a bad one, they have either not lent long enough or are not telling you.

Risk 8: concentration

One large loan in a small fund can dominate its outcome. If a fund holds $5 million and makes a $1 million loan, a total loss on that single file is 20% of the fund, and a total loss is what a badly underwritten construction loan in a falling market can deliver.

Position limits exist to convert a catastrophic outcome into a bad one. Matador Note Fund 1 LLC has a maximum committed equity of $5,000,000 and a rule that no single loan exceeds 20% of fund capital once the fund holds $20 million or more. Note what that sentence does and does not say. A smaller fund is a more concentrated fund by definition, which is a real risk of investing early in any vehicle, and it is why the size of the fund and the size of its largest loan are two of the first questions worth asking.

Concentration is not only loan-size concentration. It is also geography, asset type, borrower and exit. We lend on first-lien Texas real property only. That is a deliberate choice in favour of knowing one market and one foreclosure process deeply, and it is a concentration. Twenty of our 60 loans were to repeat borrowers, which is good evidence of underwriting and is also borrower concentration. Both statements are true at once and both are the manager's own lending record, not audited fund performance.

Risk 9: rate and refinance risk

Most hard money loans are interest-only with a balloon, with terms generally running 6 to 24 months. Our own Texas loans have averaged a 13-month term with an 11-month median, which again is the manager's lending record and not audited fund performance. Every one of those loans depends on an exit: a sale or a refinance.

Refinance risk is what happens when the exit disappears. A borrower underwrote a project assuming a long-term loan would be available at completion. Then rates rise, or the lender tightens its debt-service coverage requirement, or the appraisal comes in under the number the refinance needed. The project is finished, the borrower is not in default on anything except the maturity date, and there is no money to pay the balloon.

This is a genuinely difficult situation because the borrower is not a bad actor and the collateral is not impaired. The choices are an extension at a higher rate, a partial paydown, a sale into a market where buyers also cannot get financing, or foreclosure on a performing property. Rising rates make all four worse simultaneously, which is the definition of a correlated risk.

For the fund, the same conditions extend the average life of the portfolio just as investors are most likely to want liquidity. Loans that were supposed to repay in 12 months repay in 20. That is not a credit loss. It is still a problem.

What actually protects you

Six things, in roughly descending order of usefulness.

First-lien position. Being first in line means the borrower's equity absorbs the first losses and that a foreclosure extinguishes junior liens. Second-lien lending is a different business with a different risk profile, whatever the rate suggests.

Conservative loan-to-value. This is the whole ballgame. Our portfolio target is a 65–75% weighted loan-to-value with an 85% maximum, set by asset type: 75% target and 85% maximum on non-owner-occupied single family; 65/70 on multi-family; 65/75 on commercial; 75/80 on ground-up construction measured as-completed; and 55/65 on entitled raw land. Land is lowest for a reason. It produces no income, takes longest to sell and falls furthest.

Independent valuation. An appraisal ordered by the lender, reviewed by someone who knows the submarket, and re-ordered when a file goes sideways rather than relied on from the original file.

Title insurance and proper closing. A lender's title policy, funding through a title company, confirmed recording. This is the control that makes double-pledging and most lien surprises expensive for the fraudster and survivable for you.

Inspected draws. Money released against verified completion, not against invoices or optimism. This is the single control that most reduces the unfinished-renovation problem described above.

Position limits and real servicing. Limits cap the damage from any one mistake. Servicing determines whether a problem is found in week two or month six, and that difference is usually the difference between a full recovery and a loss.

Those controls produced our record: 55 of 60 loans repaid on schedule, five workouts, three of which returned principal in full. Two did not.

What none of it protects you from

Start with the two that did not return full principal. In both, we held a first lien, we had a valuation, we had title insurance, and we still came up short. The controls worked exactly as designed and the outcome was still a loss. Under $50,000 in total across the book is a good result and it is not zero, and the next two may not be so small.

Nothing above protects you from a market that falls further than your equity cushion. A 70% loan survives a 30% decline; it does not survive a 45% one, and no underwriting standard set today knows what the next decline will be.

Nothing protects you from time. Every control in the list slows a resolution down. Title work, notice periods, bankruptcy stays and careful workouts all cost months during which your capital earns nothing, and a loan that returns 100% of principal after 18 months of silence has still cost you 18 months of return.

Nothing protects you from being unable to get out. Lock-ups, notice, quarterly caps and liquidity gates are features of the structure, not failures of it, and they bind hardest in the conditions that make you want to use them.

Nothing protects you from the manager being wrong. Underwriting is judgement, and judgement is wrong some of the time. A record of 60 loans is enough to show a process and nowhere near enough to prove it works in a decline we have not had. Past performance is not indicative of future results, and that sentence is doing real work in this paragraph rather than sitting in a footnote.

Nothing protects you from correlation. The rate environment that stops borrowers refinancing also softens the market you would sell the collateral into, tightens the credit facility, slows payoffs and raises redemption requests. Risks in this asset class do not arrive one at a time.

And nothing protects you from a total loss on a given loan, or from losses that exceed anything in the record above. Real estate debt investments involve significant risk, including possible loss of principal. If you cannot accept that sentence as written, no amount of structure should persuade you, and I would rather you decided that now than in month nine of a workout.

Frequently asked questions

Q: What are the risks of hard money lending?
A: Borrower default, the time and cost of foreclosure, collateral worth less than underwritten, illiquidity of your position, fund-level leverage, fraud, concentration in a single loan, refinance risk when a borrower's exit disappears, and dependence on the manager's underwriting and servicing. Most of these are survivable one at a time. They tend to arrive together, because the conditions that cause one cause the others.

Q: Can you lose your principal in a debt fund?
A: Yes. A first lien reduces the chance of a total loss but does not eliminate loss. If a property sells for less than the loan balance plus foreclosure and carrying costs, the shortfall is a real loss of principal. Across our own 60 Texas loans since October 2019, total principal lost has been under $50,000, less than 0.4% of principal funded. That is the manager's lending record, not audited fund performance, and it is not a forecast.

Q: How often do hard money loans default?
A: Industry default rates vary widely by lender, asset type and vintage, and any single figure should be treated sceptically. In our own Texas record, 55 of 60 loans repaid on schedule and five went to workout: four foreclosures and one deed in lieu. Three of those five returned principal in full and two did not. That is the manager's lending record, not audited fund performance.

Q: What happens if a hard money borrower goes bankrupt?
A: A bankruptcy filing triggers an automatic stay that halts a pending foreclosure, including one scheduled for the next sale date. The lender must seek relief from the stay or wait for the case to proceed, which adds months and legal cost while the loan produces no income. A first lien generally survives the process, but the timeline stops being yours. Consult counsel on any specific case.

Q: Is private lending safer than owning rental property?
A: It is different, not simply safer. A lender sits ahead of the owner's equity and does not pay for the roof, the vacancy or the eviction, which reduces exposure to operating problems. A lender also gives up appreciation, control and the ability to sell on demand, and takes on manager risk that a direct owner does not have. Both can lose money.

If you want to go through the five workouts file by file, including the two that came up short, book a call. That conversation is more useful than any document, and it is the one I would want if I were on your side of the table.

Important disclosures. This article is for informational purposes only and is not an offer to sell or a solicitation of an offer to buy any security. Any offer is made only through the Private Placement Memorandum, Operating Agreement and Subscription Agreement of Matador Note Fund 1 LLC, which govern in all cases and should be read in full, including the risk factors. Interests are offered only to accredited investors as defined in Rule 501 of Regulation D, pursuant to Rule 506(c); accredited status must be independently verified. Real estate debt investments involve significant risk, including possible loss of principal. The 8% preferred return is non-cumulative and is payable only from available distributable cash at the manager's discretion; it is not guaranteed. The 8–12% figure is a target objective, not a promise or projection of results. Targeted returns are not guarantees of future performance, and past performance is not indicative of future results. Figures describing prior lending are the manager's own lending record and are not audited fund performance. Nothing here is legal, tax or investment advice. Matador Lending LLC is not a registered investment adviser.