Structural Inefficiencies in Private Mortgage Lending: A Quantitative Analysis of Information Asymmetry and Intermediation Costs

1. Introduction

The Canadian residential mortgage market has undergone significant structural transformation over the past decade. Following the introduction of the B-20 mortgage stress test guidelines and successive rounds of macroprudential tightening, a growing segment of borrowers has been displaced from institutional lending channels into the private mortgage market. This market — estimated at $30–35 billion in outstanding loans — operates under fundamentally different microstructural conditions than the institutional mortgage market, with implications for pricing efficiency, risk allocation, and consumer welfare.

Despite its size and growth trajectory, the private mortgage market remains poorly understood from a quantitative perspective. Academic literature on mortgage markets has overwhelmingly focused on institutional lending, securitization, and the dynamics of government-backed mortgage insurance. The private segment — comprising individual lenders, Mortgage Investment Corporations (MICs), and syndicated mortgage arrangements — has received comparatively little analytical attention. This gap is significant because the private market exhibits pricing spreads that appear to substantially exceed what standard credit risk models would predict.

This paper develops a formal spread decomposition framework for the Canadian private mortgage market. We decompose the observed interest rate spread between private and institutional mortgages into its constituent components: credit risk premium, liquidity premium, information asymmetry costs, intermediation costs, and residual profit margin. Our central finding is that intermediation and information costs account for a share of the total spread that is comparable to — and in many cases exceeds — the credit risk premium itself. This suggests that the private mortgage market is characterized by structural inefficiencies that create economic rents for intermediaries, rather than simply reflecting the higher default risk of the borrower pool.

The analysis proceeds as follows. Section 2 describes the institutional structure of the Canadian private mortgage market. Section 3 develops the formal spread decomposition model. Section 4 analyzes the information asymmetry mechanisms at work. Section 5 presents quantitative estimates for each spread component. Section 6 examines the potential for technology-driven efficiency improvements. Section 7 provides a comparative analysis with international private lending markets. Section 8 discusses policy implications. Section 9 addresses limitations, and Section 10 concludes.

Our contribution is threefold. First, we provide a rigorous decomposition of private mortgage pricing that separates risk-based compensation from friction-based costs. Second, we quantify the magnitude of information asymmetry and intermediation costs using publicly available data and industry estimates. Third, we identify the specific structural features that give rise to excess spreads, providing a framework for evaluating both regulatory interventions and technology-driven disruptions.

2. Market Structure

2.1 Overview of the Canadian Private Mortgage Market

The Canadian private mortgage market serves borrowers who cannot qualify for institutional lending through federally regulated banks, credit unions, or monoline lenders. The reasons for non-qualification are diverse: insufficient income documentation (common among self-employed borrowers), poor credit history, non-standard property types, the need for short-term bridge financing, or failure to pass the B-20 stress test. Industry estimates suggest approximately 65,000 private mortgage transactions occur annually in Ontario alone, with the national market estimated at $30–35 billion in outstanding loan volume.

Private mortgage interest rates typically range from 8% to 15%, compared to 5% to 6% for conventional institutional mortgages. This 300–1,000 basis point spread is the object of our analysis. In addition to the interest rate itself, borrowers face substantial upfront costs: broker commissions of 1–3% of the mortgage amount and lender fees of 1–3%, meaning that a borrower may pay 2–6% in fees before accounting for interest. These fees are typically deducted from the loan advance, reducing the net funds received.

2.2 Market Participants

The private mortgage market features three principal participant classes with distinct incentive structures and information sets.

Borrowers enter the private market as a last resort or for speed. They are informationally disadvantaged in the sense that they typically lack the financial sophistication to evaluate the reasonableness of the pricing they receive. Many borrowers are in financial distress or time-sensitive situations (e.g., avoiding power of sale, closing a purchase), which reduces their bargaining power and willingness to shop for competitive rates. The borrower's private information — their true ability and willingness to repay — is imperfectly observed by both brokers and lenders.

Lenders include Mortgage Investment Corporations (MICs), private lending companies, and individual investors. MICs pool capital from investors and deploy it across a portfolio of private mortgages, providing diversification but adding a layer of intermediation cost. Individual lenders range from sophisticated real estate investors managing large portfolios to passive investors seeking yield through single mortgage investments. Lenders face the standard adverse selection problem: the borrowers they see have already been rejected by institutional channels, and the information they receive about borrower quality has been filtered through the broker.

Mortgage brokers serve as the critical intermediary between borrowers and lenders. In the private market, the broker's role extends beyond simple matching: brokers package the deal, prepare documentation, present the borrower's case to potential lenders, negotiate terms, and coordinate the closing process. Brokers possess the most complete information of any market participant — they observe the borrower's full financial picture and know the lending criteria and preferences of multiple lenders. However, the broker's economic incentives create a principal-agent problem: brokers are compensated through upfront commissions that are tied to deal volume and size, not to the long-term performance of the loan. This creates incentives to close deals quickly rather than to optimize the match between borrower risk and lender risk appetite.

2.3 Information Flow Architecture

The information structure of the private mortgage market can be modeled as a sequential filtration process. Define the full borrower information set as $\Omega_B$, which includes the borrower's true financial condition, repayment capacity, property value, and future intentions. At each stage of the intermediation chain, information is both lost and potentially distorted:

$$\Omega_B \xrightarrow{\phi_1} \Omega_{Br} \xrightarrow{\phi_2} \Omega_L$$

where $\Omega_{Br}$ is the broker's information set after borrower interaction, $\Omega_L$ is the lender's information set after broker presentation, and $\phi_1, \phi_2$ are information filtration functions satisfying $\Omega_L \subseteq \Omega_{Br} \subseteq \Omega_B$.

The first filtration, $\phi_1$, represents the borrower-to-broker information transfer. Borrowers may withhold unfavorable information (strategic concealment) or fail to provide relevant information due to lack of sophistication (unintentional omission). Brokers conduct due diligence, but the depth and quality of this diligence varies significantly across the industry.

The second filtration, $\phi_2$, represents the broker-to-lender information transfer. This is perhaps the more consequential loss. Brokers prepare a "deal package" for lender consideration, which necessarily summarizes and frames the borrower's situation. The lender receives a curated presentation rather than raw data, and must make lending decisions based on this filtered view. Critical information — such as the borrower's demeanor, the broker's qualitative assessment of repayment likelihood, or nuances in property condition — may not be captured in the written package.

This sequential information filtration creates compounding losses. If each filtration preserves a fraction $\alpha_i$ of the original information content, the lender observes a fraction $\alpha_1 \cdot \alpha_2$ of the original signal. Industry estimates suggest that private mortgage underwriting relies heavily on property appraisals and title searches, with limited ability to verify borrower income or assess repayment capacity in the way that institutional lenders can through access to tax records, employment verification, and credit scoring models.

2.4 Regulatory Environment

The Financial Services Regulatory Authority of Ontario (FSRA) regulates mortgage brokers and agents under the Mortgage Brokerages, Lenders and Administrators Act. Brokers must be licensed, maintain errors and omissions insurance, and comply with disclosure requirements. However, the regulatory framework is primarily designed for the institutional mortgage market and imposes lighter requirements on the private lending transaction itself. Private lenders (other than MICs) are not subject to the same prudential regulation as banks or credit unions, and there is no standardized reporting framework for private mortgage default rates, lending volumes by risk category, or borrower outcomes.

This regulatory asymmetry means that market-level data on the private segment is sparse and unreliable. There is no equivalent of the Home Mortgage Disclosure Act (HMDA) data that enables detailed analysis of the US institutional mortgage market. The opacity of the market is itself a contributor to inefficiency, as it prevents the price discovery mechanisms that would narrow spreads in a more transparent market.

3. Spread Decomposition Model

3.1 Framework

We model the interest rate on a private mortgage as the sum of a risk-free rate and a series of premia reflecting distinct economic frictions:

$$r_{private} = r_f + \sigma_{credit} + \lambda_{liquidity} + \gamma_{info} + \delta_{intermed} + \pi$$

where:

• $r_f$ is the risk-free rate, proxied by the Government of Canada 5-year benchmark bond yield
• $\sigma_{credit}$ is the credit risk premium compensating for expected default losses
• $\lambda_{liquidity}$ is the liquidity premium reflecting the illiquid nature of private mortgages
• $\gamma_{info}$ is the information asymmetry cost arising from adverse selection, moral hazard, and search frictions
• $\delta_{intermed}$ is the intermediation cost reflecting broker commissions, administrative overhead, and transaction costs
• $\pi$ is the residual economic profit or rent captured by market participants

This decomposition builds on standard frameworks in credit pricing but introduces the explicit separation of information and intermediation costs from the credit risk premium. In efficient markets, $\gamma_{info}$ and $\delta_{intermed}$ would be minimized through competition and transparency, and $\pi$ would converge toward zero. The magnitude of these components relative to $\sigma_{credit}$ serves as a measure of market inefficiency.

3.2 Credit Risk Premium Estimation

The credit risk premium $\sigma_{credit}$ compensates the lender for expected default losses. We estimate this using the standard expected loss framework:

$$\sigma_{credit} = PD \times LGD \times \frac{1}{1 - PD \times LGD}$$

where $PD$ is the probability of default and $LGD$ is the loss given default.

Probability of Default (PD): Industry estimates place default rates in the Canadian private mortgage market at 5–15%, varying significantly with loan-to-value (LTV) ratio, property type, and borrower characteristics. For a representative portfolio with moderate LTV (65–75%), we use a baseline $PD$ estimate of 8%.

Loss Given Default (LGD): LGD in mortgage lending is determined by the shortfall between the outstanding loan balance and the net recovery from property sale. For private mortgages at 65–75% LTV, assuming property values are broadly stable and accounting for the costs of power of sale proceedings (legal fees, maintenance costs, time-to-sale discounts), we estimate $LGD$ at approximately 25–30%. This estimate accounts for the fact that private mortgages are frequently second or third liens, which increases loss severity.

The annualized expected credit loss is therefore approximately:

$$\sigma_{credit} \approx 0.08 \times 0.275 \approx 0.022 = 220 \text{ basis points}$$

However, this expected loss must be adjusted for the term structure of default. Private mortgages are typically 1-year terms (compared to 5-year terms for institutional mortgages), which compresses the default window. On an annualized basis, for a 1-year term, the expected credit loss is:

$$\sigma_{credit}^{annual} = PD_{1yr} \times LGD \approx 0.06 \times 0.275 \approx 165 \text{ basis points}$$

where $PD_{1yr}$ is the 1-year default probability, which we estimate at approximately 6% for a representative portfolio. We adopt a central estimate of $\sigma_{credit} \approx 150\text{–}200$ basis points.

3.3 Liquidity Premium Estimation

Private mortgages are highly illiquid instruments. Unlike residential mortgage-backed securities (RMBS) or commercial mortgage-backed securities (CMBS), there is no secondary market for individual private mortgages. A lender who wishes to exit a position before maturity faces substantial costs: negotiating a loan sale to another private lender, potentially at a discount to par; arranging an assignment; or waiting for maturity.

To estimate the liquidity premium, we compare to instruments with similar credit characteristics but different liquidity profiles. Canadian CMBS and high-yield corporate bonds provide relevant benchmarks. The liquidity component of CMBS spreads has been estimated at 50–150 basis points in various market conditions. For private mortgages, which are substantially less liquid than any traded instrument, we estimate a liquidity premium of approximately $\lambda_{liquidity} \approx 75\text{–}125$ basis points, with a central estimate of 100 basis points.

3.4 Intermediation Cost

The intermediation cost $\delta_{intermed}$ is the most directly observable component. It includes:

• Broker commissions: 1–3% of the mortgage amount, typically 2% on average. For a 1-year mortgage term, this translates directly to 200 basis points annualized. For shorter terms or renewals, the annualized cost is even higher.
• Lender fees: 1–3% of the mortgage amount, also typically deducted from the advance. Average lender fee of 1.5% on a 1-year term equates to 150 basis points annualized.
• Legal costs: Borrowers typically pay for both their own and the lender's legal counsel, adding $2,000–$5,000 to transaction costs.
• Appraisal and due diligence: $300–$1,000 for property appraisal, plus title insurance and search costs.

On a $300,000 private mortgage (a representative amount), total upfront fees of 3.5$10,500, or 350 basis points annualized for a 1-year term. Including legal and due diligence costs, the total intermediation cost is approximately:

$$\delta_{intermed} \approx 200\text{–}250 \text{ basis points (annualized for 1-year term)}$$

This estimate represents the directly observable intermediation cost. Note that some intermediation costs are embedded in the interest rate rather than charged as upfront fees, which means our estimate may be conservative.

3.5 Total Spread Decomposition

Using a representative private mortgage rate of 10% and a Government of Canada 5-year bond yield of 3.5%, the gross spread is:

$$S_{gross} = r_{private} - r_f = 10.0$$

We can now decompose this spread:

The residual component — encompassing information asymmetry costs and economic profit — ranges from 75 to 225 basis points. This residual is large relative to the credit risk premium itself, suggesting significant pricing inefficiency.

Key finding: Intermediation costs (200–250 bps) are as large as or larger than credit risk costs (150–200 bps). In an efficient market with transparent pricing and low-friction matching, we would expect intermediation costs to be a small fraction of the credit risk premium, not a comparable magnitude.

3.6 Sensitivity Analysis

The spread decomposition is sensitive to several key parameters. We examine the sensitivity of the residual to variations in inputs:

Scenario A: Low default, low LTV (60%): $PD_{1yr} = 4$, $LGD = 20$, $\sigma_{credit} = 80$ bps, Residual = 195–345 bps

Scenario B: High default, high LTV (80%): $PD_{1yr} = 12$, $LGD = 40$, $\sigma_{credit} = 480$ bps, Residual = negative (rates may exceed 10% for this segment)

Scenario C: High-rate private mortgage (14%): $S_{gross} = 1050$ bps, with moderate risk: Residual = 475–625 bps

The sensitivity analysis confirms that the residual is robustly positive for moderate-risk private mortgages (the bulk of the market) and becomes very large for higher-rate transactions where intermediation costs are proportionally similar but the rate spread is wider.

4. Information Asymmetry Analysis

4.1 Adverse Selection

The private mortgage market exhibits classic adverse selection as described by Akerlof (1970). Borrowers who enter the private market have been filtered through the institutional lending process and rejected. From the lender's perspective, the private borrower pool is adversely selected: it disproportionately contains borrowers whose risk characteristics exceed institutional thresholds.

However, the adverse selection in private mortgages is more nuanced than a simple "lemons" problem. The institutional rejection criteria include both hard factors (credit score, debt ratios, stress test failure) and soft factors (income documentation, property type, non-standard situations). Many borrowers in the private market are not inherently high-risk; they simply do not fit the standardized underwriting templates of institutional lenders. Self-employed borrowers with strong businesses but non-standard income documentation are a canonical example.

We can model this using a two-type framework. Let $\theta \in {H, L}$ denote borrower quality (high or low). In the institutional market, the screening technology $S_I$ correctly identifies and rejects a fraction $\beta$ of low-quality borrowers and incorrectly rejects a fraction $\alpha$ of high-quality borrowers (Type I error). The private market borrower pool then has a quality composition:

$$P(\theta = H | \text{private}) = \frac{\alpha \cdot q}{\alpha \cdot q + (1-\beta)(1-q) + (1-q)\beta_{\text{miss}}}$$

where $q$ is the population fraction of high-quality borrowers and $\beta_{\text{miss}}$ accounts for low-quality borrowers who also failed institutional screening. If the institutional screening has even modest Type I error ($\alpha > 0$), the private market contains a non-trivial fraction of high-quality borrowers who are overcharged relative to their true risk.

The key insight is that private market lenders, lacking the screening technology of institutional lenders, cannot efficiently distinguish between the genuinely high-risk borrowers and the "false rejects" from the institutional channel. This inability to sort drives pooling pricing: all private borrowers pay rates that reflect the average risk of the pool, creating a cross-subsidy from lower-risk to higher-risk borrowers.

4.2 Moral Hazard

Post-origination moral hazard arises from the limited monitoring capacity of private lenders. Unlike institutional lenders who have dedicated servicing departments, automated payment tracking, and established workout procedures, many private lenders — particularly individual investors — have minimal infrastructure for ongoing loan monitoring.

The moral hazard operates on two dimensions:

Borrower moral hazard: Once funded, borrowers may take actions that increase default risk — deferring property maintenance, taking on additional debt, or diverting funds intended for mortgage payments. In the institutional market, these risks are mitigated by covenant structures, automated monitoring of payment behavior, and cross-default provisions linked to other credit facilities. Private mortgages typically lack these safeguards.

Broker moral hazard: Since brokers are compensated at origination, they bear no ongoing economic exposure to loan performance. This creates an incentive to originate deals that may not be in the long-term interest of either borrower or lender. The broker's informational advantage — having observed the borrower's full financial picture — means they may be aware of risk factors that are not fully communicated to the lender. While regulatory requirements mandate fair dealing, the practical enforcement of these requirements in the private market is limited.

The moral hazard cost can be modeled as a monitoring premium:

$$\gamma_{MH} = c_m + (1 - p_m) \times \Delta PD \times LGD$$

where $c_m$ is the cost of monitoring, $p_m$ is the probability that monitoring detects moral hazard, and $\Delta PD$ is the incremental default probability attributable to unmonitored moral hazard. For private mortgages with minimal monitoring infrastructure, $p_m$ is low, and the moral hazard premium is substantial.

4.3 Search Friction

Perhaps the most underappreciated source of inefficiency in the private mortgage market is search friction. The process of matching a borrower with an appropriate lender is manual, relationship-driven, and time-consuming. Average time to fund a private mortgage is 2–6 weeks, compared to 3–5 days for institutional mortgages.

The search friction arises because:

1. Lender heterogeneity: Private lenders have diverse and idiosyncratic risk appetites. Some specialize in urban properties, others in rural; some accept higher LTV, others require lower; some lend on commercial properties, others only residential. There is no standardized classification system for lender preferences.

2. Bilateral negotiation: Each deal is individually negotiated. Unlike the institutional market, where rates are posted and largely standardized for given risk categories, private mortgage terms emerge from bilateral bargaining between broker and lender.

3. Relationship dependence: Brokers match borrowers to lenders based on existing relationships. A broker's "lender network" is their primary competitive asset, but this means the matching is constrained by the broker's network rather than the full universe of available capital.

4. Sequential search: Brokers typically approach lenders sequentially rather than running a simultaneous auction. If the first lender declines, the deal moves to the next, with each step adding time and potentially leaking information about the deal's rejection history.

The cost of search friction can be estimated from the time value of delayed funding. If the average private mortgage funds 3 weeks later than an institutional alternative, and the borrower has an alternative cost of capital of 15% (typical for borrowers in distress), the search friction cost is approximately:

$$\gamma_{search} = r_{alt} \times \frac{\Delta t}{52} \approx 0.15 \times \frac{3}{52} \approx 0.87$$

This is borne by the borrower as a cost of market participation, though it does not appear directly in the interest rate. It represents a deadweight loss attributable to market structure.

4.4 Combined Information Cost

Combining the three information asymmetry channels — adverse selection, moral hazard, and search friction — we can estimate the total information cost component of the spread:

$$\gamma_{info} = \gamma_{AS} + \gamma_{MH} + \gamma_{search}$$

Based on our analysis:

• Adverse selection pooling cost: 50–100 basis points (estimated from the cross-subsidy between false rejects and true high-risk borrowers)
• Moral hazard / monitoring premium: 25–75 basis points
• Search friction (embedded in rate): 25–50 basis points

Total information cost: $\gamma_{info} \approx 100\text{–}225$ basis points

This estimate is consistent with the residual from our spread decomposition in Section 3, providing cross-validation of the framework.

5. Quantitative Estimates and Empirical Calibration

5.1 Baseline Calibration

We calibrate the spread decomposition model to a representative private mortgage with the following characteristics:

• Mortgage amount: $300,000
• Property value: $450,000 (LTV: 67%)
• Location: Greater Toronto Area, urban residential
• Borrower: Self-employed, limited income documentation
• Term: 12 months, interest-only
• Private mortgage rate: 10.0%
• Comparable institutional rate: 5.5%
• Government of Canada 5-year bond yield: 3.5%

Under these parameters, the gross spread of 650 basis points decomposes as follows:

5.2 Key Finding: Intermediation Dominance

The most striking result is that intermediation costs (34.6% of the total spread) exceed credit risk costs (26.9%). When combined with information costs (15.4%), the non-risk friction components account for 50% of the total spread. This means that half of what private mortgage borrowers pay above the risk-free rate compensates for market structure inefficiencies rather than the actual risk of lending to them.

To contextualize this finding, consider the institutional mortgage market. For a conventional insured mortgage, the spread over the risk-free rate is approximately 150–200 basis points, of which credit risk accounts for the majority (supported by mortgage insurance pricing, which reflects actuarial default estimates). Intermediation costs are minimal — perhaps 10–30 basis points — because the institutional market benefits from standardized processes, automated underwriting, and scale economies. The private market's intermediation cost is roughly 10 times that of the institutional market on a per-dollar basis.

5.3 Fee Structure Analysis

The fee structure of private mortgages introduces a non-linearity that exacerbates the cost to borrowers. Because broker and lender fees are charged as a percentage of the mortgage amount but the mortgage term is typically 12 months, the effective annualized cost of these fees is:

$$r_{effective} = r_{stated} + \frac{f_{broker} + f_{lender}}{T}$$

where $T$ is the term in years. For a 12-month term:

$$r_{effective} = 10$$

For a 6-month bridge loan at the same stated rate:

$$r_{effective} = 10$$

This term-dependence of effective cost creates a perverse outcome: the borrowers most in need of short-term bridge financing face the highest annualized costs, precisely because the fixed intermediation fees are amortized over a shorter period. This is a structural feature of fee-based intermediation that would be mitigated in a more efficient market structure with volume-based or outcome-based compensation.

5.4 Portfolio-Level Analysis

At the portfolio level, a private mortgage portfolio with the above characteristics would exhibit the following return decomposition for the lender:

• Gross yield: 10.0%
• Less: expected credit losses: –1.75%
• Less: servicing and administration: –0.50%
• Less: broker commissions (if lender-paid): –2.00%
• Net yield to lender: 5.75%

Compare this to a 5-year GIC rate of approximately 4.0–4.5%. The private mortgage lender earns a net premium of approximately 125–175 basis points over a risk-free alternative, after accounting for credit losses and intermediation costs. This modest net premium — for an illiquid, operationally intensive investment — raises questions about the risk-adjusted attractiveness of private mortgage lending for individual investors, particularly those who lack portfolio diversification.

For MICs, the economics are different. MICs charge management fees (typically 1–2% of assets under management) and may retain a portion of broker fees, creating additional layers of intermediation cost that are ultimately borne by the MIC investors through reduced net returns and by borrowers through higher rates.

6. Technology Disruption Potential

6.1 Sources of Technological Efficiency

The intermediation and information costs identified in our analysis suggest significant potential for technology-driven efficiency improvements. We identify four primary channels through which technology could compress the non-risk components of the private mortgage spread.

Automated document verification and underwriting: The current private mortgage underwriting process requires 2–10 hours of manual work per deal. This includes reviewing borrower-submitted documents, verifying property details, assessing title, and preparing the deal package. Automated document processing — including optical character recognition, bank statement analysis, and property data integration — could reduce this to 30–60 minutes. Estimated cost compression: broker and lender staff time could be reduced by 60–80%, translating to 50–100 basis point reduction in intermediation costs.

Algorithmic lender-borrower matching: The current relationship-driven matching process is the primary source of search friction. A platform-based matching system that maintains a database of lender preferences and automatically routes qualifying deals to appropriate lenders could reduce the matching process from days or weeks to hours. This would also enable a form of competitive pricing: presenting the deal to multiple lenders simultaneously rather than sequentially, introducing price competition that currently does not exist for most private mortgage transactions. Estimated search friction reduction: 50–70 basis points.

Enhanced risk assessment: Automated property valuation models (AVMs), integrated credit data, and algorithmic risk scoring could reduce the information asymmetry between the three market participants. By providing lenders with standardized, data-driven risk assessments rather than broker-curated deal packages, the adverse selection and information filtration problems could be substantially mitigated. Estimated information cost reduction: 30–60 basis points.

Standardized deal structures: The bilateral, bespoke nature of private mortgage transactions creates legal and administrative overhead. Standardized loan documents, automated closing processes, and digital title registration could reduce per-transaction costs. Estimated cost saving: 20–40 basis points.

6.2 Aggregate Efficiency Gains

Combining these channels, the total potential efficiency gain is estimated at:

Under this estimate, technology-enabled intermediation could compress the combined intermediation and information cost from approximately 325 basis points (225 + 100) to approximately 100–150 basis points. This would reduce the representative private mortgage rate from 10.0% to approximately 8.0–8.5%, holding all else equal.

6.3 Barriers to Disruption

Several factors may limit the pace and magnitude of technology-driven efficiency gains:

1. Regulatory requirements: Anti-money laundering (AML), know-your-client (KYC), and suitability obligations require human judgment that may resist full automation. Regulatory frameworks may need to evolve to accommodate automated processes.

2. Data availability: Effective algorithmic underwriting and risk assessment depend on historical data. The private mortgage market's opacity means that training data for machine learning models is scarce, creating a cold-start problem.

3. Relationship entrenchment: Incumbent brokers and lenders have built their businesses on relationship networks. The value of these relationships creates switching costs and may generate resistance to disintermediation.

4. Trust and adoption: Both borrowers (often in vulnerable financial situations) and lenders (deploying significant capital) may be reluctant to rely on automated processes for high-stakes transactions. Building trust in technology-enabled platforms takes time.

5. Market fragmentation: The private lending market is highly fragmented, with thousands of individual lenders and hundreds of brokers. Achieving the network effects necessary for a platform-based model requires critical mass on both sides of the market.

7. Comparative Analysis

7.1 International Private Lending Markets

The structural features of the Canadian private mortgage market can be contextualized through comparison with analogous markets in other jurisdictions.

United States — Hard Money Lending: The US hard money lending market is the closest analogue to Canadian private mortgages. US hard money rates typically range from 10% to 18%, with origination fees of 2–5 points. Funding speed is faster than in Canada — often 7–14 days — reflecting a more mature market infrastructure with established lending platforms and more standardized practices. The US market benefits from larger scale (estimated at over $100 billion outstanding), which supports more sophisticated secondary market activity and data availability. Several technology platforms have successfully entered the US hard money space, compressing intermediation costs for transactions that fit standardized parameters.

United Kingdom — Bridging Finance: The UK bridging loan market serves a similar function, with rates of 0.5–1.5% per month (approximately 6–18% annualized). The UK market is more regulated than the Canadian private market, with the Financial Conduct Authority (FCA) providing oversight for regulated bridging loans. Industry bodies have developed standardized disclosure requirements and lending criteria, which have improved price transparency. Intermediation costs, while still significant, have been compressed by online platforms and increased competition.

Australia — Non-Bank Lending: Australia's non-bank lending sector has evolved differently from Canada's. Non-bank lenders in Australia often operate at institutional scale, accessing wholesale funding through securitization. This means that Australian non-bank mortgage rates are typically 50–200 basis points above bank rates — dramatically lower than the 300–1,000 basis point premium observed in Canada. The difference is attributable to Australia's more developed non-bank securitization market, which provides scale, liquidity, and standardization that the Canadian private market lacks.

7.2 Comparative Spread Decomposition

The comparative analysis reveals that Canada's private mortgage market has the highest intermediation share among the markets examined. This is consistent with the market's relative immaturity, lack of standardization, and limited technology adoption. The Australian non-bank market, which has the lowest intermediation share, demonstrates what is achievable when private lending achieves institutional scale and standardization.

7.3 Maturity Assessment

We can rank private lending markets by structural maturity using several indicators:

1. Secondary market development: Australia > US > UK > Canada
2. Technology platform adoption: US > UK > Australia > Canada
3. Standardization of terms: UK > Australia > US > Canada
4. Regulatory framework: UK > Australia > Canada > US
5. Data transparency: US (HMDA data) > UK > Australia > Canada

Canada's private mortgage market ranks last or near-last on most maturity indicators, consistent with our finding that non-risk spread components are disproportionately large. The market appears to be in an earlier stage of structural development than its international peers, suggesting significant potential for efficiency improvement as the market matures.

8. Policy Implications

8.1 Transparency Requirements

Our analysis suggests that enhanced transparency requirements could address some of the information asymmetry that drives excess spreads. Specific policy measures could include:

Standardized rate disclosure: Requiring that private mortgage rates and fees be disclosed in a standardized format (similar to the Annual Percentage Rate (APR) disclosure for institutional mortgages) would enable borrowers to compare offerings and improve price discovery. Currently, the combination of stated interest rates, broker fees, lender fees, and legal costs makes it difficult for borrowers to compare the true cost of alternative offerings.

Market-level data collection: Requiring private lenders (or brokers) to report aggregate data on lending volumes, rates, default rates, and borrower characteristics would address the data gap that prevents systematic market analysis. FSRA or a designated statistical agency could collect and publish this data in anonymized form, analogous to US HMDA reporting.

Broker compensation disclosure: While brokers are required to disclose their fees to borrowers, the disclosure is often buried in complex documentation. Requiring prominent, plain-language disclosure of all intermediation costs — and how they compare to industry benchmarks — would empower borrowers to negotiate more effectively.

8.2 Market Infrastructure

Beyond transparency, structural reforms could reduce intermediation costs directly:

Centralized deal matching: A regulatory-supported or industry-sponsored platform for matching borrowers with lenders could reduce search frictions. Such a platform need not replace brokers entirely but could supplement the relationship-based matching process with data-driven alternatives.

Standardized documentation: Developing standard-form private mortgage documents would reduce legal costs and processing time, analogous to the standardized documentation used in the institutional securitization market.

Quality certification: A voluntary or mandatory quality certification system for private lenders could reduce adverse selection on the lender side (borrowers choosing among lenders of unknown quality) and increase trust in the market.

8.3 Regulatory Calibration

Regulators face a delicate balancing act. The private mortgage market serves a legitimate and important function: providing credit to borrowers who cannot access institutional channels. Excessive regulation could shrink the market, forcing these borrowers into even more costly alternatives (unsecured credit, predatory lending) or leaving them without access to credit. The goal should be to improve market functioning — reducing inefficiency without reducing access.

Our spread decomposition provides a quantitative framework for evaluating regulatory interventions. Policies that reduce $\gamma_{info}$ and $\delta_{intermed}$ without increasing $\sigma_{credit}$ (i.e., without tightening lending standards to the point of excluding borrowers) would improve market efficiency. Policies that primarily tighten supply — such as imposing institutional-style capital requirements on private lenders — risk raising rates for borrowers by reducing competition among lenders.

9. Limitations

This analysis is subject to several important limitations that should be considered when interpreting the results.

Data constraints: The private mortgage market lacks systematic data collection. Our estimates of default rates, transaction volumes, and market size are based on industry surveys, regulatory reports, and practitioner estimates rather than comprehensive administrative data. The true values may differ from our estimates, and the uncertainty ranges we provide reflect this data limitation.

Model simplification: The spread decomposition model treats each component as additive and independent. In practice, there may be interactions between components. For example, higher information asymmetry may lead to higher default rates (if lenders systematically misprice risk), creating a feedback loop between $\gamma_{info}$ and $\sigma_{credit}$. Our additive framework does not capture these second-order effects.

Geographic specificity: Our analysis focuses primarily on the Ontario market, particularly the Greater Toronto Area. Private mortgage market conditions vary significantly across Canadian provinces and between urban and rural markets. Spreads and intermediation costs in smaller markets may be higher due to thinner lender participation, while some markets may exhibit different structural features entirely.

Static analysis: Our decomposition is a point-in-time estimate. The private mortgage market is dynamic, with spreads, default rates, and competitive conditions changing with macroeconomic conditions, housing market cycles, and regulatory changes. A dynamic model that captures how the spread decomposition evolves over the interest rate cycle would be valuable but is beyond the scope of this paper.

Selection effects: Our representative mortgage (67% LTV, GTA urban residential, self-employed borrower) may not be representative of the full distribution of private mortgage transactions. The spread decomposition will differ for higher-LTV loans, rural properties, construction financing, and other market segments.

Unobserved heterogeneity: The residual in our spread decomposition — which we attribute to information costs and profit — may include other unobserved components. For example, operational risk costs (the risk of fraud, documentation errors, or property valuation errors) may be significant in the private market and are not separately identified in our framework.

Despite these limitations, we believe the qualitative conclusions are robust: the private mortgage market exhibits substantial non-risk spread components, intermediation costs are a major contributor to borrower costs, and there is significant potential for efficiency improvement.

10. Conclusion

This paper has presented a quantitative analysis of the Canadian private mortgage market through the lens of spread decomposition. Our central finding is that the interest rate spread between private and institutional mortgages — typically 400 to 1,000 basis points — is not primarily a function of credit risk. Instead, intermediation costs and information asymmetry account for approximately half of the total spread in a representative private mortgage transaction.

The decomposition reveals that intermediation costs (200–250 basis points) are comparable in magnitude to credit risk costs (150–200 basis points), a finding that is robust across a range of parameter assumptions. Information costs — arising from adverse selection in the borrower pool, moral hazard due to limited monitoring, and search frictions in the matching process — contribute an additional 75–150 basis points. The remaining residual, which we interpret as economic profit, is modest (25–75 basis points), suggesting that the excess spread is primarily attributable to market structure rather than monopoly rents.

These findings have implications for multiple stakeholders. For borrowers, they suggest that a significant portion of the cost of private mortgage credit is not an inevitable consequence of their risk profile but rather a function of market structure that could, in principle, be reduced. For lenders, the findings highlight that net risk-adjusted returns in private mortgage lending are substantially lower than gross yields suggest, once expected credit losses and intermediation costs are properly accounted for. For regulators, the spread decomposition provides a framework for evaluating which interventions would improve efficiency (e.g., transparency requirements, standardized documentation) versus those that would primarily restrict supply (e.g., institutional-style capital requirements).

The comparative analysis with international markets reinforces these conclusions. Canada's private mortgage market is structurally less mature than comparable markets in the United States, United Kingdom, and Australia, with higher intermediation shares and less developed market infrastructure. The Australian experience, where non-bank lending has achieved institutional scale through securitization and standardization, demonstrates that dramatic compression of intermediation costs is achievable.

Technology-enabled disruption offers the most promising path toward efficiency improvement. We estimate that automated document processing, algorithmic lender-borrower matching, enhanced risk assessment, and deal standardization could compress combined intermediation and information costs from approximately 325 basis points to 100–150 basis points. However, realizing these gains requires overcoming significant barriers, including data scarcity, regulatory adaptation, and the entrenchment of relationship-based intermediation.

The private mortgage market's excess spread represents an efficiency gap, not merely a risk premium. Closing this gap would benefit borrowers through lower costs, lenders through better risk-adjusted returns (via improved risk assessment), and the broader financial system through more transparent and efficient credit allocation. The path toward closing this gap will likely involve a combination of regulatory evolution, technology adoption, and market maturation — processes that are already underway but that the quantitative framework developed here can help accelerate and evaluate.

References

Akerlof, G. A. (1970). The market for "lemons": Quality uncertainty and the market mechanism. The Quarterly Journal of Economics, 84(3), 488–500.

Stiglitz, J. E., & Weiss, A. (1981). Credit rationing in markets with imperfect information. The American Economic Review, 71(3), 393–410.