As much as crypto offers a futuristic vision of a world monetary system, it is not a safe investment. The implosion of the FTX exchange this past November exposed the inadequate financial reporting of the crypto ecosystem, among other weaknesses. On the other hand, agency Mortgage-Backed Securities (MBS) – backed by real property – is an established asset class regulated by the Financial Industry Regulatory Authority (FINRA). In fact, an addendum to the Trade Reporting and Compliance Engine (TRACE) rules proposes to shorten the timeframe for reporting MBS trades from the current fifteen minutes to under one minute, ensuring the integrity of the reporting.
Such regulation is just one dimension when it comes to the stability of agency MBS. When we compare the one-year returns of the Bitcoin ETF (BITO) versus those of MBS ETF (MBB) and S&P 500 (^GSPC) (Exhibit 1), we observe that Bitcoin returns have been severely more volatile and plummeting. On January 3, 2023, crypto returns stood at a whopping -64%. This is more than six times as much loss as MBS total returns at around -11%. Even in the era of Quantitative Tightening, where mortgage rates have reached the highest level in two decades, MBS investors are still able to preserve most of their capital versus crypto investors who cannot.
In addition, there is room for disruptive technology to help MBS investors find alpha. For market participants who have an affinity to fintech innovation but want to invest in a long-term asset class, a great alternative to crypto would be to invest in MBS using cutting-edge technology such as Infima’s tools powered by deep learning.
Exhibit 1: Crypto’s volatile profile still leads to plummeting returns
Sources: Infima Technologies, Market Watch
Predicated on decentralization, crypto has the allure of becoming an alternative to fiat currency. However, at any moment, the government can – and has the incentive to – suspend private cryptocurrencies.
Agency MBS, on the contrary, carries either an implicit or explicit guarantee from the U.S. government. The Government Sponsored Enterprises (GSEs) – including Fannie Mae and Freddie Mac – have been taken into conservatorship since 2009, while Ginnie Mae provides the full-faith-and-credit guarantee of the U.S. Government. The three Agencies perform duties as trustees for MBS, responsible for the distribution of cash flows and monthly data reporting.
The Federal Reserve (Fed) has also been heavily involved in Agency MBS since 2008, often holding as much as a third of all MBS. During the latest round of Quantitative Easing (QE4), the Fed again appointed Agency MBS as an instrument to stabilize liquidity in the financial system. This trust speaks to the critical importance the Fed places on supporting the mortgage market through Agency MBS.
To help investors make an informed decision, Infima’s coverage is extensive. It spans across more than half a million MBS pools securitized by the three Agencies as well as tens of millions of mortgages with guarantees.
Besides lacking a government guarantee, crypto has little history to show that it can weather the next recession. Bitcoin, for one, was first released as open-source software in 2009, following the last Great Financial Crisis. Due to crypto’s highly volatile nature, it will likely be hit harder than even equities in the next economic downturn.
Mortgages have been the engine of the economy and have proved to be some of the most reliable investments for centuries. And MBS as a structured product was first issued in 1968 and has therefore also been around for nearly six decades. Infima’s model is trained on more than a billion samples of monthly loan-level data, incorporating the track record for MBS since January 2000. This history covers a few business cycles, from the Dot-Com Bubble to the Great Financial Crisis of 2008.
Since credit standards tightened post-Crisis, lending from the three Agencies has focused primarily on good credit with full documentation, and defaults have been minimal as a result. Even in cases of delinquencies or defaults, the Agencies pledge to buy such loans out of the pool, transforming the credit loss into prepayment. Especially within the last year, when most Agency MBS have been trading at a steep discount to par, lower credit characteristics led to more prepayment and consequently contributed to higher returns.
Infima’s state-of-the-art neural networks help identify these lower credit characteristics. Its experienced model picks up complexities among credit attributes in nuanced manners that traditional models tend to miss. As shown in Exhibit 2 below, the impact of lower credit scores (FICO) rises sharply when the loan-to-value ratio (LTV) passes the 80% threshold. There are two potential rationales for this interactive and nonlinear relationship. The first reason may be that borrowers whose LTV is above 80% are incentivized to rid themselves of the mandatory mortgage insurance, and the second reason could be that more leveraged borrowers would receive higher returns on their equity when they prepay. At any rate, such credit analysis on MBS collateral is worth exploring for investors bracing for the next credit crunch.
Exhibit 2: Interactive relationship between FICO and LTV ratio on prepayment rates
Source: Infima Technologies, Inc.
Notes: This chart is produced using a pool-level analysis that ranks attributes by their influence on prepayment speeds. By including multiple attributes in the same chart, such as above, the Engine explains how these attributes jointly influence speeds, highlighting their interactivity.
At the end of the day, crypto’s value depends solely on people’s belief in it. Most cryptocurrencies have no corresponding assets or returns and, as a result, have questionable fundamental value. Conversely, Agency MBS is secured by mortgages, which are collateralized by real property. It’s hard to argue that there are many other assets more fundamental than our homes.
Even the biggest idiosyncratic risk that affects Agency MBS cashflows - prepayments - could be minimized by using predictive technology, such as the AI-powered tools from Infima. Given that investors are willing to pay more for MBS with favorable characteristics – called stories – they can leverage Infima’s toolbox to sift through them.
A low loan balance story (LB085) offers both extension and call protection, for instance, by stipulating that the original size of each loan in the pool be under $85 thousand. In either a premium or a discount environment, borrowers with smaller loan balances have fewer absolute dollars to gain or lose and are therefore less rate sensitive, all else equal. To obtain such lower prepayment risk that the LB085 story offers, investors are willing to pay the most in Agency MBS. And to hedge against this higher price risk, they would benefit immensely from getting its speeds right ahead of time.
Infima helps investors predict prepayment speeds with the highest accuracy among the incumbent data vendors. Infima’s forecasted speeds for LB085 almost mirror its realized speeds, achieving an actual-to-projected Conditional Prepayment Rate (CPR) ratio close to 1.00 (Exhibit 3). By uncovering hidden patterns in borrower behavior through its deep neural network, Infima’s model delivers reliable, unbiased projections on prepayments for most stories (Exhibit 3).
Exhibit 3: Actual-to-predicted CPR ratios by story cohort hugs the 1.00 line
Source: Infima Technologies, Inc.
Notes: These actual-to-predicted ratios are computed on a weighted-average basis using the current balance of outstanding pools to aggregate them to the story-cohort level. The underlying data include actual and Infima’s projected 1-month CPRs from January 2019 to September 2022. A ratio of 1.00 represents a “perfect foresight” model, to which Infima’s model is getting close.
With uncertainty on the horizon, more investors will be looking for safe products instead of rolling the dice on crypto. Government guarantee, track record, and fundamental values are only several of the reasons investors should consider the safety of Agency MBS.
And one more – if the world ever enters a crisis where the power grid is compromised – crypto, running on lots of electricity, would lose all its value. Tangible assets such as real estate, however, would more likely survive.
To learn how to find safe bets in MBS that most suit your investment needs, check out Infima’s series on selecting attractive MBS pools.