Launch HN: Lofty AI (YC S19) – Real estate investment with alternative data

[loftyai]

Hi Hacker News Community,

My name is Jerry, and I’m one co-founders for Lofty AI (https://www.lofty.ai/). We use machine learning to help identify homes where values are likely to appreciate, and we help home buyers buy them. People can partner up with us to buy a recommended property. If they do, we are willing to cover any potential losses on the property. In exchange, the buyer agrees to share some of the future profit on the home with us. The agreement lasts 3 years.

Before starting this company, my co-founder and I had tried to invest in homes. However, we quickly got tired of realtors telling us to make offers based on very little data. We wanted to figure out a way to buy affordable homes that had the highest growth potential via a data driven approach. We realized there was a wealth of new alternative data out there, which could be used to predict both neighborhood growth and individual property growth. This alternative data we envisioned ranged from the growth in the number of postings on social media about a specific dog breed, to the number of restaurants in an area serving a specific type of trendy food, to the average wait time for ride sharing apps, and the average maximum temperature an area can experience.

Our tech involves running clustering to identify trends and keywords from text based data (e.g.: social media photo tags, business reviews) that are associated with different categories of neighborhoods (for example: rich/suburban/static, middle-class/urban/growing). We then take these insights and feed them into a larger model with historical home prices, house level features, and an array of other numeric features (e.g. ride sharing wait times, new businesses) that predicts future home price on both an individual property and neighborhood level. With this trained model we can then predict future home prices based on these alternative data sources (as well a few traditional data sources). As we ingest more data going forward we are constantly retraining and reoptimizing our models. Along with successful backtesting we have been tracking our predictions to validate our models in production and have found that properties we had identified 12 months ago have beaten the market in appreciation by an average of 14 points (yay!).

Most young working professionals want to live in or near large metropolitan cities for the lifestyle and better jobs market. This has contributed to extremely high home prices for places like the bay area and many young professionals end up paying rent that is on par with a mortgage payment. However, instead of building equity in their own future through an investment, they are simply making their landlords richer.

We want to change this by giving people another option. They can now invest in a home and their capital can be protected should the investment flop. The trade off is that these homes tend to be located in areas not “currently” deemed to be a desirable neighborhood. In essence, we want to help inexperienced home buyers make smarter decisions, and we are willing to risk our own capital for that. In the event of a downturn in the market we are hedging our exposure by buying deep out of the money options that track the real estate market. These hedges are also attached to each individual contract so even if we were to go out of business before the maturation of the agreement or before a downturn in the market your downside protection would still be alive and well! As a result, anything that’s above a 20% decline across the portfolio will be covered by the hedging instruments, so we only need to be able to guarantee the range between 0 to -20% using our own capital. To make sure we can abide by the guarantee, we know exactly how many contracts we can enter into, and we will not go above that threshold until we obtain more funding.

Sign up with us to receive a list of recommended properties that our models think will appreciate over the next 3 years. Make an offer on the property you like the most using any method you’d like. If you don’t have an agent you work with, we can recommend you one along with helping you get a mortgage. After you make an offer on a home, you enter into a contract with us. We agree to cover losses over the next 3 years and in exchange, you share some of the future upside with us.

Let us know if you have any questions or insights, and I’ll be happy to respond! Feel free to directly reach out to me at [email protected] as well. We’d love to hear your feedback and suggestions!

[onlyrealcuzzo]

I'll start by saying that I assume you know much more about the market than me, given that you've started this company and made it into YC.

If I read your post right -- the way your insurance works is:

I'm a home buyer. I think the housing market is frothy right now, but I want to buy a home anyway. So I can use your insurance to protect myself in the event the value of my house decreases in the future.

Your company is directly responsible for the first 20% decline. After that, other companies are responsible for the rest. And even if you go out of business, I'm still covered by those other companies.

Firstly, is that correct?

Secondly, if so -- what happens if you go out of business and my house goes down exactly 20%? I assume that means I have no coverage.

Thirdly, who's insuring the houses after a 20% decline? And why should I have any reason to believe they'd still be in business if the market collapses 20%+? The last time that happened, almost every insurance company and investment bank went out of business.

Finally, how much does this cost as a percentage of the house's current value yearly? Roughly...

It's an interesting idea. Despite the fact that most home buyers anticipate house price appreciation to underperform historical averages (and a lot of buyers actually think prices will go down) -- a lot of houses are being bought. I think a lot of those people would want an option like this.

[loftyai]

Thanks for your question!

I believe my main post or the responses might have been unclear. If so, my apologies.

But your understanding isn't correct. Other companies are not insuring your downside. We are the only counter party you have.

The problem is if a recession happens, then a lot of our properties actually decline in value. As a result, we might not be able to pay you back. So to make sure we can pay you back we buy financial instruments on the open market, kind of like buying a stock of apple for example. These instruments work in a very interesting way. Their prices go up, if the real estate market goes down. Their prices go down, if the real estate market goes up.

So, with these instruments. We can ensure that in the event of a recession, we can still afford to pay you back, because we can sell the instruments for higher prices than we originally paid for. We then use that profit to cover the losses our customers experience.

The way this works out is that events that would cause large declines in the property values are covered by these instruments. Which means, as a company, we just need to pay specific attention to the potential losses between 0-20% range. Here, we deposit the 20% value of the original purchase price into the 3rd party account.

In the event that our company stops operation. These hedging instruments don't expire or disappear. They are bought at the beginning of our agreement with our customer. As a result, these instruments will be passed off to our lawyers along with the 3rd party account for them to maintain. This way, your loss coverage will still be guaranteed even if we go out of business.

Is this more clear? If not, I can always elaborate :)

[findjashua]

my guess is that they're buying put options on the shiller home-price index, at a strike price 20% below the spot price.

though, those indices are only granular at the city level, whereas during a recession all neighborhoods in a city don't drop in value by the same rate - eg for bay area in 2008, east bay got decimated, whereas palo alto/peninsula barely dropped 5%.

[whoisjuan]

Do you have any type of financial insurance to backup of your claim that you will cover losses if the property sells for less? I'm not talking about having the money available to cover the losses, but actually being around at all to honor that claim.

What happens if I buy today and your company goes to hell in two years? How can I trust this transaction with a horizon of 3 years without fully knowing how are you going to perform as a company.

Of course, if you disappear and I don't have to honor the 20% it's a win for me, but if I buy based on your data and after 3 years the property value is way down and you're not around to honor your loss-covering promise then I'm fucked. I bought a house because you told me it was going to appreciate, but it didn't. Now what? Do you have a way to pay me back even if your company is not around anymore?

There's something about the model that doesn't make sense to me.

[loftyai]

Great question! So, we actually maintain a 3rd party account that is only allowed to invest in short-term US treasury notes.

We track all the properties in our portfolio daily. Any on paper depreciation will result in us depositing funds into the 3rd party account. Whenever a property price moves above the original purchase price on paper, we will withdraw any previously deposited fund. This on-going process along with the hedging instruments are what allows us to guarantee the downside protection.

As a final layer of protection, we know exactly what our on going exposure is, so we know the maximum amount of contracts we can underwrite. We are very strict on this number and will never move above it. So, even if our company ceases operations, all of the downside protection will still be available to our customers.

Keep in mind, we also know exactly what our on going

[Hydraulix989]

My first reaction was: If this company goes under (like 90% of YC startups) and you took this sort of arrangement, then you are @#$!'d (all caps).

[Jemaclus]

For years the simplest heuristic I can find for finding up-and-coming places has been "Where is Starbucks opening new stores?" I assume Lofty's is much more complicated than that, but I'd be curious to see what the overlap is between the Starbucks Strategy (TM) and Lofty AI's.

I do have a few more questions though:

- Also, are you focusing on primary residences, homes as investments (i.e., rentals)? Do you consider apartments, duplexes, or commercial real state?

- Do you have an idea of how long one would have to own these homes for the appreciation to appreciate in a significant enough way for it to be profitable?

- You mention "some of the future profit". How much is that? 1%, 5%, 10%, 50%?

[loftyai]

Great question! The Starbucks Strategy is actually well known in the industry. Believe or not, large real estate developers and investors will often follow the same signals. They also look for things like Trader Joe's or Whole Foods opening.

Our algorithm is very similar in concept to this strategy. However, by the time Starbucks or Trader Joe's opens in an area, it's often towards the middle or late stages of a neighborhood's growth. We can find amenities that are even earlier indicators than Starbucks. Think your one-off local coffee shop named "Bob's coffee" or something similar.

We are focused on the appreciation potential of residential real estate, which has single family houses, condos, and town homes. However, we have noticed that in areas where home prices are growing, rents typically are growing as well. So our customers are welcome to rent out the properties for cash-flow.

We do not have data on a lot of commercial properties, but we can still underwrite the agreement on duplexes and smaller multifamily units.

It typically takes 3-5 years on average for neighborhoods to see the exponential portion of their growth curve, so our agreement is for 3 years by default.

Our share of the profit is 20% of the gross profit. So, if you had bought something for 100,000 and you sold it for 200,000 in 3 years. Then, we would get 20% of the gross profit ($100,000), which would be $20,000.

edit: made numbers in example more clear.

[malloreon]

according to the site, you keep 80% of the profit, so their cut is 20%.

[SimianLogic2]

Why did you decide on this business model (essentially downside insurance paid for by equity if I'm understanding it right) over something simpler like, say, subscription access to a newsletter?

What's to stop someone from signing up for the list and just buying a property on their own? (i.e. What perks are you offering that make it worth doing the deal through you? Negotiations? Acting as a buyer's agent?)

As a data point, I used to have a ruby script that would take a bunch of MLS IDs and go pull a ton of facts from Zillow and a few other sources. I would have my realtor set up a high-level search (i.e. SFH in these areas under $500k) and then take their daily emails and run them through my script to identify potentially "undervalued" properties. I still had to hand-check them after, but it was a pretty useful second filter (the MLS search being the first).

[loftyai]

Thanks for your question! We decided to do this model, because we originally sold our predictions and analytics to larger investment funds, but we noticed that when our predictions came true, we left so much money on the table. The funds were making millions of dollars on one deal and they were never going to give us any percentage of that.

It was also really hard to convince a lot of these people who were operating on "gut feelings". In January of this year, we made the prediction that Compton, LA was going to see an increase in growth. We told these bigger funds and they literally laughed at us during the meeting. Fast forward to today, and some of the properties in the micro-neighborhood we forecasted showed an 18% growth in price in just 7 months.

So we decided that consumers might find what we are building to be more valuable, and they would be more open to sharing the profit with us if our predictions came true.

Our added benefit is really finding neighborhoods that people overlook, but have high growth potential. Realistically, without our platform, I would have never known about the growth or be interested in Compton, LA either.

Right now, there is a paywall to view the listings. It's $100/month, but you may cancel at any point. Additionally, if you end up signing a contract with us, we refund you all the money you've paid up to that point. If people do not do the contract with us, then they would also not be offered the downside protection.

I love hearing about people's own unique technical method for finding properties! Were you able to invest in any properties using your method?

[ttcbj]

This is the first time I have ever seen an early stage company include Saint Louis in anything, so, thanks for that ;-).

That said, you say your market is:

"Lofty AI is best for people who are: 1. Thinking of buying their first home, but are nervous about losing money. 2. Looking for higher returns than normal by buying properties in an appreciating neighborhood early."

1. I wonder if people who know they have to sell in the next three years, but don't want to sell today (e.g. a work move) are also a target market. If I know my job is going to move me in 2 years, I might like to use your service to retain 80% of the upside, but insure against downside when I sell.

2. I once read a book about real-estate investing, which said that the real way you make money is to buy rental properties with poor cash flow, 'fixup' the tenants to improve the cash flow, and then sell, repeatedly. I wonder if your appreciation-potential-evaluation/downside-insurance model applied to rental properties for sale, combined with coaching/tools for aspiring landlords, might be attractive.

It seems like right now, you are primarily using the purchaser as a source of capital, and other comments are saying "why don't you just raise the money yourself?", but if you were also using the purchaser as more like a franchisee, someone who is actively working to improve the cashflow of the property by upgrading the tenants with your (automated) advice, that might create a more interesting relationship where you have more room to add value (its more complex to analyze multi-tenant rentals, its more complex to choose high-potential landlord partners, etc).

Random thoughts. It's a very interesting idea, very original.

[loftyai]

Thanks for the feedback, you bring up some really interesting points.

1. I think what you're saying (and please correct me if i'm wrong) is that we could go after people who aren't about to buy a home but who already own a home and may want to sell it in 2-3 years. That is actually something we already have done and are open to doing more! We could certainly make it more explicit on our website that this is an option.

2. This is a really cool concept. As you have noted we aren't so much in the business of encouraging people to optimize the cash flow on a home and partnering with them on that, but this is a common way to make money of real estate and is certainly something we could branch out into.

This would complement our goal of not having to just become a fund and help solve real pain points people have in purchasing homes really well.

Again, really appreciate this feedback - the phrasing sparked some really cool insight and will definitely think about this more going forward.

[tryitnow]

Do you have any customer testimonials? It would be good to link to them.

This is definitely something I would never consider doing unless I've heard other people doing it, I would never want to be the guinea pig here.

And honestly, if I am financially clever enough to understand your value proposition then I'm probably financially clever enough enough to buy some downside protection on general real estate assets.

The most interesting value proposition here is the ability to predict future home prices, but if you could really do that you would be working for one of the massive real estate funds, for the same reason that someone who is really good at picking stocks will work for a hedge fund (and eventually create their own), that person isn't going to decide to give investment advice to a gazillion pipsqueak investors or manage the accounts of a bunch of little investors, even if they can easily automate it.

In other words, the ability to predict asset prices is something that it makes sense to keep as private as possible, not something to share with the masses.

[loftyai]

Customer testimonials is a great idea! Will add those to our site when we get a minute to do so.

As you said, they would probably go along way towards providing some additional comfort to any one who has some interest but is cautious about moving forward.

The questions you raise regarding our business model are good ones. As mentioned elsewhere in the comments, a fund is something that would be interesting but that we just don't have the capital for at the moment. Furthermore, our initial motivation for creating Lofty was to address the pain point of people wanting to buy a home but being cautious about the risk. As such a pivot to a fund, while similar in nature (and perhaps simpler in some ways), would be a pivot from addressing a real pain point we see in the market to just becoming another real estate fund and is in part why we are hesitant to do so, on top of the higher capital requirements.

Appreciate the feedback!

[eaenki]

Since your model consistently beats the market and you can hedge at city level, Wouldn’t you be better off just raising money for an hedge fund that goes long and short with some leverage?

It’s a cool idea anyway, good luck

[loftyai]

It's certainly something we have considered.

One issue is the capital raise. On top of that we really wanted to address the specific pain point of people wanting to buy a home but who cannot afford making a bad purchase. As such a pivot to a fund model would be a pivot from our initial motivation in starting Lofty. Nonetheless, it's an interesting idea.

Appreciate the feedback!

[kevinguh]

You mention that you hedge your exposure to market downturns through deep OOTM options -- would it be safe to interpret this as your company taking out OOTM puts on various REITs/ETFs? If so, I'm wondering about a couple things:

1. Do you hedge on REITs/ETFs with a local presence in the areas your properties are located in? If so, is there any liquidation risk of the REIT/ETF in the event of a major downturn that could force an early exit from your hedge position and leave you exposed to further decline? Also, how would you handle rebalancing/constructing new hedges when you add investment properties in a new area?

2. If you hedge on broader diversified REITs/ETFs, is it a plausible concern that your investment properties can be hit by a localized recession that leaves other parts of the broader real estate market unaffected, thus leaving your hedge unable to recoup the losses?

[loftyai]

Yes, your interpretation is correct!

1. we hedge on both broader market REITs/ETFs as well as localized ones, depending on how many contracts we have in the local market.

2. Because we hedge on both, the probability of this is very low. Since a more granular hedge is an imperfect hedge due to the nature of these REITs/ETFs, it might not cover 100% of the localized recession. However, it should cover a large portion of it. So, our company will be on the hook for that remainder percentage.

We can cover it in 2 ways. Number 1, just use our own capital. Number 2, the profitable contracts in other areas not hit by recessions should be able to offset the ones hit by the localized recession.

[josephpmay]

Against my better judgment, I'm gonna comment that I hate this.

This is certainly a good business/investment opportunity. If your algorithms are any good you'll make a lot of money, and you'll help your customers make money.

My problem with Lofty is that it is bad for society.

Fundamentally, this is gentrification-as-a-service. You're driving additional demand to neighborhoods at inflection points, and if it works well it's definitely going to accelerate displacement. It is true that Blackrock et al already do this and likely have similar in-house algorithms, but this will widen the market and make things worse. I'm not completely against gentrification, especially when there's development that increases market supply- and when done right that can actually decrease displacement. (an example you're probably familiar with is the USC Village. Despite the criticism it gets, it removed thousands of student renters from the South Central LA market which likely resulted in downward pressure on prices) But housing speculation like this drives up prices and only hurts poor and minority communities.

To Jerry and Max: did you consider the ethical implications when deciding to start this? I completely see the angle that you're democratizing access to an investment asset that currently mainly benefits wealthy institutional investors. I imagine, though, that a lot of people are going to see your team of young almost-all-white males and paint you as everything that's wrong with tech, and you should consider to what degree that assumption represents the truth.

[loftyai]

You're absolutely right about this, and it is something we have and are considering everyday. I will be honest and say, at the moment, we do not have a perfect solution yet.

One of the things we looked into before starting the company was a paper that mentioned the Portland project, which showed that gentrification and displacement are not always synonymous. There, the neighborhood was completely gentrified, but the locals benefited greatly, because many of their home prices increased in value, and many owned local businesses that benefited from the influx with affluent people.

One of our goals is to see how we can use our data and business to make gentrification more like the Portland project. One idea has been to provide our data and analytics to city governments for free, so they can act faster in regards to setting up affordable housing.

In the meantime, our customers will be buying the homes, so someone has to act as the seller. If a local resident was the owner of the property, then hopefully, they benefit from the sale (we recommend our customers offer the "listing price" and not negotiate at all). If they are the renter, then current California laws should provide them a lot of protection.

It's not perfect of course, but we are looking for better alternatives.

[fapjacks]

These are precisely my thoughts, too. I try to keep myself from falling prey to the ultimate human indulgence of claiming to live in some kind of moral decline preceding some end times, but this is exactly the kind of thing that scares the shit out of me. Once people are paying 40% or 50% of their income to rents because they have no choice, what comes next? Combined with all of the enormous consumer debt...

[syntaxing]

This is a super interesting idea, and I have been wondering if there was a service like this. Is there something similar to this that recommends the house based on your stage of life? For instance, if you have a kid and your income is X amount, this is probably your best area or even if I wanted to retire with a set standard of living and maybe environment. This paired with home value appreciation that you provide would be awesome!

[loftyai]

Love this suggestion! We have been toying with the best way to personalize the properties shown to each individual user. Right now it's just by city, max price and property type.

We also have added things like "near nightlife", "near trendy coffee spots", "near schools" in an attempt to capture what you have suggested here - albeit in a less efficient manner.

A "stage of life" questionnaire would be a great way to encapsulate all those above and more in an easy to interface with UI for the user. Thank you!!

[jpn]

Are you hedging with REIT options? Or home builders?

Are you able to hedge specific city risk?

[loftyai]

Yes, we actually hedge with both depending on the area the property is located in. I hope you don't mind me not mentioning the specific products, since we don't want people in the market to bid up our hedging instruments and make them more expensive for us to buy.

We can look into the portfolio for a lot of the REITs as well as the exposure certain builders have. Based on that, we can hedge specific states and cities very well. It's not a perfect hedge, but it will definitely reflect the local real estate market.

[jedberg]

Pretty cool concept! You're basically letting people buy into your real estate hedge fund. :)

Something I didn't see addressed in the FAQ: Do I get to include closing costs and agent fees in my cost basis before determining your 20% cut?

If not I can see a case where appreciation was minimal and I actually lose money one you get your cut because those fees eat up all the appreciation.

[loftyai]

Thanks for the comment, but at the moment we do not cover closing costs or agent fees.

In the event that there is appreciation, but is minimal. We recommend that our client buys us out instead of selling the property. This way, they won't have to deal with the transactional costs, and with the low appreciation, the 20% cut will be very low, so it shouldn't be hard to come out of pocket for that.

As I've mentioned in some of the other replies, we'd love to keep innovating in this space and help reduce these fees, so our clients no longer have to deal with them.

Open Door is one of the companies doing a lot of cool stuff in this space, but we think there can be a lot of room for improvement. Statistically, a home is roughly 64% of the average American's total lifetime net worth, and we don't think people should have to pay so much fees on top of handling their life's most valuable physical asset.

[rajacombinator]

It’s an odd business model. I think if you had any faith in your predictive ability you would just raise and run the fund yourself so as not to leave money on the table. This sounds like all the “crypto trading as a service” scams that have come up over the past few years. YC should avoid getting involved in trading businesses they don’t understand. Just my 2c.

[loftyai]

This is totally understandable reaction to the business model. We do put our money where out mouth is but nonetheless it does beg the question of why we don't just do the investments ourselves.

One major issue is capital. But beyond that we also see a saw a market need for a product like this when trying to make home purchase decisions in personal life prior to creating Lofty. And that was the market and problem we wanted to address with this company - not just to be a big fund that buys real estate en masse and profits from it - but to help people who can't afford to make a bad home investment do so more comfortably.

[deepnotderp]

Have you considered measuring RF activity in areas to approximate the number of people? I'd also imagine that a higher ratio of telecom bands as compared to wifi bands would indicate more bias visitors and therefore potentially more future growth.

[loftyai]

This is a great idea! We do not currently do this, but we do take into consideration the availability of gigabit internet as well as overall connection speeds for different neighborhoods.

This has actually proven to be statistically significant for a large number of locations across the united states.

[Fluent_Startup]

Very cool. I just forwarded to a friend looking to buy. As a real estate investor, I can see why this would benefit a lot of people. I personally buy from a handful of turnkey operators (Midsouth Homebuyers has a 3 month vacancy guarantee), but may consider this in the future.

[alephnan]

> Before starting this company, my co-founder and I had tried to invest in homes.

Were you guys successful?

[loftyai]

Haha not at all, since we never ended up investing. We were inexperienced and didn't really know what to look for. Homes that required extensive rehab seemed too daunting and a turnkey property in a nice neighborhood was too expensive.

(This will sound like we are very lazy...) We essentially wanted to buy an affordable home that would just grow in price over the next few years without us doing anything. At the time, there were no tools to help us find homes like this, which ultimately led us to starting this company.

[rog211]

Do you use all public data like census, BLS etc. or do you have paid data as well?

[loftyai]

We do not use census data, because they are extremely outdated. It's part of the reason we can make accurate predictions before other companies can, since they do look at census data.

We use alternative data, which has recently become popular in the finance industry. For example, if you ask executives at a big company what their profit outlook is, they will always be optimistic, otherwise, their stock might decline and they may panic the market.

If you waited until the quarterly announcement, then you would be finding out at the same time as everyone else, and it's delayed information.

However, some people have found that you can more accurately predict a company's outlook on their quarterly performance by monitoring job boards and see how many open positions the company is hiring for. This allows people to gain insight and act before the rest of the market catches on.

We use the same approach but for real estate. For example, if you monitor the number of french bull dogs in a neighborhood, you can accurate predict median income values for that neighborhood before any official statistics. This is because those dogs are very expensive, so someone willing and able to spend a few thousand dollars on a pet tend to have a higher economic background.

We do use some paid sources such as satellite imagery and some data sources require you to pay for their api like our weather data vendor.

[mayank]

Are there any case studies you'd like to share, either here or on your site, or any other validation of your methodology? At the moment, it looks a bit like a "hot stock picks" newsletter.

[loftyai]

We do have case studies, but they are bad in the case of presentation. We are a small team so when we decided to keep our analytics internal facing, we didn't spend anytime producing marketing materials or prospectuses.

So, most of the case studies are loss values printed onto our engineer's console or .png graphs showing our walk forward predictions outputted from our engineer's notebook.

If you'd like I can dig through our slack channel to find some stuff for you. I'll check back in later to see If I can find something more presentable as well.

[loftyai]

If you're still interested, you can email me directly, and I'd be happy to send you some of the stuff I was able to dig up.

[RosanaAnaDana]

Hey, this seems pretty cool. What are your thoughts on landscaping/ outdoor water use?

If that seems interesting to you and your team, you should hit me up.

[meerab]

What is role of Real estate agent in your business model? Why would a buyer be willing to shell another 3% for a Buyers Agent?

[loftyai]

This might not be the case everywhere in the country, but typically, the seller is responsible for the agent fees. This means that if we recommend an agent for our customer, who is the buyer, that agent is actually paid for by the seller.

We have the relationship, because we do have customers who are very inexperienced and this would be their first purchase. So, a lot of them still want to have to ability to talk to an agent and ask questions about the home buying process.