Tuesday, September 13, 2016

Off Point: The Case for Home Loans

UPDATE: I had a nice discussion with Alex Rampell, partner at VC firm a16z, who provided me with a material detail that I missed that is both positive and negative. I greatly appreciate his time and while it does change the value proposition for homeowners, it remains unclear to me that the new benefit outweighs the new cost for most homeowners. That said, with the new information I can see where a homeowner with a sizable equity stake could value the flexibility it may provide and I do see where this could become a viable product a few iterations down the line.

What did I miss?

I missed that Point does in fact share in the downside with the homeowner if the home price falls below Point's "risk-adjusted value" (and the loss has not eaten away all of the homeowners equity). Point places a risk-adjusted value on each home at the time they take a stake (i.e. if the home has a market value of $1,000,000, it sounds like they'll typically strike a risk-adjusted value at some price below that). Once the home declines below that risk-adjusted price, Point shares in the loss at their ownership stake (i.e. if they took a 10% equity stake in the home, they share in 10% of the losses below that risk-adjusted value).

I would note that the risk-adjusted value is also an incremental cost to the homeowner that I initially missed. If a $1,000,000 home has a Point risk-adjusted price of $900,000 and is sold the very next day at $1,000,000, then it sounds like Point is entitled to a 20% cut of the $100,000 "gain" even though the market didn't move.

Where I personally can see this eventually being a viable product is for an individual that bought low (likely a long time ago), is now sitting on a house worth a LOT of money, and is liquidity constrained (i.e. cannot afford the monthly interest payments associated with a HELOC). Retirees planning to move in the next ten years (Point requires they are paid out of their equity stake within ten years) would be one demographic falling into this camp, as would technology folk who have been buying up property all over San Francisco when we hit our next tech recession (please please please happen sooner than later).

My dialogue with Alex also made it pretty clear to me that the longer term goal of Point is to build a platform where all these details would more dynamically sort themselves out. For now, the initial analysis below does understate the downside (i.e. the amount captured by Point is actually higher given the risk-adjusted price), but also understates the upside (i.e. the homeowner can benefit by more than the interest savings if prices move lower). For those with a high LTV (loan-to-value), thus lower equity cushion, I believe the initial analysis remains a good enough proxy for the payout structure as the benefit of the put received (i.e. the loss Point offsets in a market decline) is only there until the total equity is wiped out (banks get paid back first).


I am pretty sure there will be a ton of other posts tomorrow about Point (a "home equity partner") given the huge response to Marc Andreessen's tweet (if not... check out the responses here - UPDATE: the great Matt Levine has a take here). As a result, I'll skip whether or not the fractional home equity investor is really just a predatory lender, how they will / won't disrupt the mortgage industry, or whether it marks the VC top. Instead I'll focus on the investment case against swapping out of a traditional mortgage loan and "selling" Point a fraction of your home equity. I make a lot of assumptions in the below, but if anything looks off, please let me know. TLDR: don't do it.

Swapping from a Loan: A Homeowner's Perspective

The real benefit of fractional equity for homeowners will be for those that cannot afford the initial 20% down payment and cannot borrow from family. Unfortunately, Point will not solve this issue as:

To be eligible for Point, you’ll need to retain at least 20% of the equity in your home after Point's investment.
As a result, this is a pretty straight forward swap for a homeowner that is otherwise reliant on borrowing. The homeowner is simply swapping out the cost of interest payments for the opportunity cost of the housing appreciation associated with whatever equity stake is sold should markets rise.

In the below analysis, I assume a $1,000,000 home where an investor swaps out 10% ($100,000) from a ten year interest only loan and instead "sells" that 10% stake to Point.

Assuming a 3% interest only loan for ten years and a 33% tax rate for interest deduction benefits, we get a 2% / year after-tax savings (3% x [1 - 0.33%] = 2%) for the homeowner due to the reduced notional loan. This comes out to $2000 of savings per year or $20,000 over a ten year period. In return, the homeowners gives up 200% of the upside on the stake sold and is charged 3% up front. Unfortunately, the homeowner doesn't appear to receive any incremental benefit should markets fall as the homeowner takes first loss on the minimum 20% equity stake homeowners are required to retain. Andreessen's VC firm a16z outlines:
Point gets paid back after the bank, but before the homeowner,
Given Point gets paid back before the homeowner, this capital structure provides a huge benefit to Point at the expense of the homeowner (i.e. heads Point wins, tails the homeowner loses). There could be examples where Point also takes a hit, but the homeowner appears to have already been wiped out an amount equal to Point's stake at that point.

Payout Diagram

The below payout diagram assumes an investor doesn't pay back the loan for the max 10 years (the longest period of time before Point is owed their investment back) and that the increase in home value comes solely from market appreciation (i.e. no upgrades paid for by the homeowner). The dotted line represents the starting house price and we can see that the homeowner is better off if the price does not move or if the price declines, as they pocket the $20,000 they would have otherwise paid in after-tax interest. In return, the homeowners break even if the market rises a bit more than 1% a year, given the interest savings and 3-4% up front cost, a rate that is forecast to be less than the rate of inflation.

Unfortunately it gets worse if the homeowner remodels (one of the benefits of owning a home). Per point:
Many homeowners use Point funds to pay for remodeling costs. You can remodel your home but there are limits on adding additional debt to the home to pay for remodeling. Please also note that Point will share in any increase in property value due to remodeling changes you might make.
Let's be clear... if a homeowner sells a $100,000 stake in order to invest in a new kitchen, increasing the value of the home by $100,000, Point captures 20% of that $100,000. Note that $20,000 ($100,000 x 20%) equals the entire benefit of avoiding the interest payments for the full ten years, meaning there is no longer any net upside of choosing Point vs a loan at time = 0. The dotted line represents the new starting point assuming the $100,000 was used to upgrade the home (note the green line cross at ~$0).

As for the cost at inception, these payoff diagrams are classic naked call payoffs and the interest savings can be viewed as the premium collected for selling those calls (i.e. the $20,000). Assuming these calls expire in 10 years, we can try to value them using the Black-Scholes formula. Using the assumed inputs:
  • Notional: $200,000 (200% upside of $100,000)
  • Strike: $1,000,000
  • Current (Remodeled Home) Spot: $1,100,000 
  • Risk free rate (assumed 2%)
  • Volatility (the annual volatility from 2006-2016 for the Case Shiller index was 9%)
We get a value of roughly $60,000. Assuming any home price movement will be on top of 2% / year inflation, these calls move up to a rough value of $100,000. All for $20,000 in saving.

Keep the loan.

Tuesday, September 6, 2016

How a Low VIX Can Remain an Expensive Hedge

One of my favorite Twitter follows @LadyFOHF shared the below scatter chart from Morgan Stanley that attempted to map areas of the global market that were both cheap (valuation ranks at the lower end of its 10-year history) and defensive (a low or negative correlation to global equities).

One of the few trades listed as having both characteristics was the VIX Index. Let's take a look.

The VIX is Not Investable

The chart below shows the percentile rank of the VIX index over the ten year time frame outlined in the Morgan Stanley chart. The VIX Index was indeed near it's all time low, ranking in the lowest 6th percentile of its ten year history at Friday's close (9/2/16) when the VIX closed at 11.98. 

But an investor cannot invest in the VIX Index. Rather an investor can buy or sell the consensus view of where the VIX will be, most easily through VIX futures or ETPs (for more details on the VIX term structure, take a look at a past post here).

VIX Futures have Diverged

The below looks at the rank of historical 2nd month VIX futures contract prices (i.e. where consensus views the VIX as being when the 2nd closest VIX future contract matures - which currently makes up the largest holding of VIX ETPs) and we see a different picture. As of Friday's close, the second month contract was "only" in the 30th percentile in terms of price at a value of 16.375, more than 35% higher than the VIX Index itself.

VIX Futures Rank Divergence vs Future Returns

This 20%+ gap in rank of the VIX Index and the 2nd month VIX futures contract is pretty extreme, indicating investors believe (and are pricing in) a higher VIX in the future that is more "normal" to history; a very different view than the exuberance the current level of the VIX implies.

Going back to the initial question of whether the VIX is cheap... the below chart takes the historical gap (from the green chart above) and projects out forward daily performance (annualized) of the S&P 500® VIX Short-Term Futures Index (the index benchmarked within the VXX ETP) of various scenarios. Similar gaps show that VIX prices were at extremely expensive levels (not cheap levels) and it actually presented a buying opportunity for the S&P 500 (11.9% returns when the gap was positive). 

To summarize... while an investor should hardly ever be long VIX futures as they rarely provide a positive return, it is the the term structure of the VIX (i.e. VIX futures relative to the current level of the VIX), rather than the level of the VIX itself, that is likely to be a better signal of whether VIX futures are rich or cheap.