Gal Krubiner

The through-line

Everything Krubiner says returns to a single idea: the point of applying artificial intelligence to consumer credit is to let lenders say yes more often. He frames Pagaya's mission as being "to provide access to credit more often to people" [2][3], and he arrives at consumer credit not because he set out to fix lending but because he was looking for the market where big data and AI could do the most damage to the status quo fastest. "Before even we got to the consumer credit I think the premise was big Big Data AI which are buzzwords today but back then were tools to assess risk," he says, and consumer credit turned out to be "one of the areas that has the atmost amount of data and has the ability to make big changes rather quickly" [2][3].

The second constant is a deliberate refusal to compete with the people he serves. Having concluded that the first wave of online lenders "did such a great job call it fintech lending 1.0 that we as another competitor not going to bring a real value to the world," he chose the enabling seat: "we went to the back seat of the B2B b2c model and is an enabler of at the beginning fintech lenders today just every lender or a bank out there" [2][3]. The arc from 2016 onward is visible in how he describes the company: first as "an asset manager in a rapidly growing asset class of online credit" whose 2017 job was "execution and about acquiring more assets to manage" and team building [1], later as a two-sided AI lending network running "fully automated 247" behind fintechs, banks, credit unions and auto dealers [2][3].

On the 42% who get declined

Krubiner's framing of the problem is both economic and human. Despite four decades of progress in the American financial system, he argues, "still there is something like 42% of people that are getting rejected decline whatever you want to call them in the moment when they are applying for a credit" [2][3]. He insists on reading that two ways at once. There is the efficiency loss, since the applicant "already came to the place they already look for the credit" and is turned away anyway. And there is what he calls the emotional piece: the person who now "needs to look for a different solution and has and is experiencing a bad day because of the" decline notice [2][3]. That second reading is what makes the decline population a mission rather than a market segment for him, and it recurs when he describes borrowers pushed out of mainstream credit by rate caps as "a negative impact on millions of lives of people in the US" [2][3].

On the mechanics of partnership

Pagaya's method is embedding rather than referring. The company connects by technology into partners and puts "our AI and capabilities into the loan origination systems of these different partners," allowing them to approve borrowers they had decided to decline, with a lift of up to 20% in better approvals and more funding in good cases [2][3]. The commercial hook, in his telling, is that the lender keeps the relationship while Pagaya takes the risk: "while this consumer is becoming their consumer and recognizing their brand we are still coming back with the ability to fund it and to take it off their balance sheet" [2][3].

He is candid that the two halves of the model are currently welded together. Underwriting and investment are correlated, so a partner cannot today take the models without the balance sheet, though he says Pagaya is "thinking about opening that and provide it as service" for partners who want the underwriting capability while keeping the assets themselves [2][3]. The partner mix he points to spans personal loans, auto and point of sale: Westlake routing consumers from its full dealership networks, a bank he identifies as the biggest subprime lender bank in the US, and Klarna in POS, alongside institutional capital from GIC, the sovereign wealth fund of Singapore, plus names such as Angelo Gordon [2][3]. He also cites a partnership with one of the world's largest asset managers to provide $100 million plus of capital to a credit union [2][3]. The public discussion of this investor–lender–borrower network as a way of reshaping credit underwriting is the substance of his Wharton FinTech appearance as well [4].

On why more data means more approvals

Krubiner is explicit that Pagaya's edge is positional rather than exotic. Operating in a heavily regulated market means being "very much bounded to the FCRA compliant data," so the inputs are mostly credit bureau data plus proprietary data [2][3]. The differentiator is the vantage point: "while A monoline lender has its own data set when we are sitting in the middle and having this unique point of view of the different flow and the different changes that all the lenders are doing it's informing us" [2][3]. He meets the obvious objection, why a bank would hand over that data, with an alignment argument rather than a defensive one: "as long as we get more data and the data sets is more robust we can by definition approve more consumers," so partners collaborate to feed the models the most robust data possible [2][3].

Model evolution follows the same logic of widening the aperture. The progression ran from straightforward personal loans into auto and then point of sale, producing a view of the consumer "from different type of angles that usually a specific lender is looking on it only from the angle they are looking to provide a credit" [2][3]. He describes the effect geometrically, as "a box that is every time getting a little bit bigger and a little bit bigger and a little bit bigger," where yesterday's outskirts become today's mainstream [2][3].

On pricing as a cause of default, not just a consequence

One of his more distinctive arguments is that price is an input to risk rather than only an output of it. Beyond market competition and adverse or positive selection, he points out that the rate you set determines the monthly payment, "and therefore will impact the probability of default of that borrower to pay" [2][3]. Pricing a loan at 12, 14 or 16 percent changes the borrower's ability to stand within the payment. What Pagaya's research department has learned over time, he says, is to treat dynamic pricing as a factor that belongs inside the approval decision, framed around "what is the thing that will be most helpful for them from a lending perspective" [2][3].

On the ABS market and investor demand

Krubiner treats Pagaya's securitisation shelf as proof of institutional acceptance. He describes the personal loan ABS shelf as the dominant one in the US and argues that "today for many investors if you want to get exposure to Consumer Credit you're actually choosing pagaya" because it is among the largest, most liquid and best known [2][3]. On the demand cycle, he dates the low point precisely to the end of Q4 2022, with demand strengthening through 2023 as talk of a better economy took hold, as the consumer proved able to push the economy, and as the weaker performance of the vintages originated at the height of 2021 and the start of 2022 faded from view. His evidence is that transactions come "over subscribed one after the other in the market" [2][3].

On the state of the consumer and the credit crunch

He separates two phenomena that are often confused. Consumer payment behaviour is stable, with no sign of deterioration in ability to pay, particularly for anything originated in the last year, and he judges the inflation wave of 2021 and 2022 to be "kind of like behind us from a consumer perspective," with the caveat that oil prices could change that [2][3]. Credit availability is the opposite story: the steepest and longest decline he has seen, running consistently from Q3 of the prior year and with headwinds strengthening. He attributes it to two causes, Fed rate policy and the banking crisis, specifically the lack of liquidity at mid-size banks, which he calls "by definition the strongest force for providing this type of loans and liquidity to the market" [2][3].

Rising rates hit two populations hardest, at opposite ends. Subprime, lower-FICO borrowers who a year earlier would have priced under the 30 or 36 percent APR cap are now above it on a one-for-one rate increase, "and therefore they are out from the mainstream ability to borrow from a regular lender" [2][3]. At the other end, the 2.99 percent car financing offer no longer exists; with a 5 percent floor plus spread, 8 percent becomes the floor even for super prime borrowers, so many simply use savings or cash rather than borrow [2][3]. The middle sees somewhat more expensive credit and less of it. He does not soften the conclusion: the subprime exclusion is a negative outcome, "sometimes necessary to tame the inflation" but damaging all the same [2][3].

On why tight credit is a tailwind for Pagaya

Asked directly whether scarce credit hurts a business built on expanding credit, he says the reverse: less available credit means "we are filling the gaps," and it acts as a catalyst for the business model [2][3]. He puts the shift in historical terms. Where 2008 pushed the big banks toward conservatism, he argues that the 2023 events and Silicon Valley Bank in particular changed how mid-size banks, super regionals and credit unions think about being competitive and pricing aggressively [2][3]. The consequence he expects is structural: fintech adjusting "a little bit more to Capital markets Asset Management money rather than depositary capital," a shift that favours a player with strong footing on both sides [2][3]. His scoreboard argument is blunt: lending businesses in this environment typically shrink 20 or 30 percent while Pagaya keeps growing, so "the proof is in the pudding" [2][3].

On building the company

The earliest milestone he singles out is not a product launch but a mandate. Onboarding the first client after more than a year of work, and being given discretionary authority to manage their money, was the defining moment of 2016, and he describes it as the point at which prospective clients saw the value the team was trying to create [1]. He positioned the firm from the start as a finance-oriented company leveraging AI and technology in the online credit space, working with big institutions and private money alike [1], and set 2017 as a year "focused all about execution," acquiring more assets to manage and building out the team across functions [1].

Location was a strategic decision of the same kind. He moved to the States in 2018 because "it was very clear that New York is the capital of the capital of the world," and because the people making the right decisions sit inside those buildings, so he should be among them [2][3].

Takeaways

• Roughly 42% of credit applicants are declined at the moment of applying, a figure Krubiner treats as both an efficiency failure and a personal one for the applicant [2][3].
• Pagaya deliberately avoided competing with fintech lenders, taking "the back seat of the B2B b2c model" as an enabler for lenders and banks instead [2][3].
• The models sit inside partners' loan origination systems and can produce up to a 20% lift in approvals, with the loans taken off the lender's balance sheet while the lender keeps the customer relationship [2][3].
• Data is mostly FCRA-compliant bureau data; the edge comes from sitting in the middle of many lenders' flows, and more data mechanically means more approvals, which aligns partners' interests with Pagaya's [2][3].
• Pricing should be treated as an input to default risk, since the rate set determines the monthly payment and therefore the borrower's probability of paying [2][3].
• Consumer ability to pay has been stable while credit availability has seen its longest decline, driven by Fed policy and mid-size bank liquidity after SVB [2][3].
• A 5% fed funds floor pushes subprime borrowers above the 30 to 36 percent APR cap and out of mainstream lending, while super prime borrowers increasingly pay cash rather than borrow at 8 percent [2][3].
• Tight credit is a tailwind for the business, and Krubiner expects fintech to lean more on capital markets and asset management money than on depositary capital [2][3].