All Forum Posts by: Stefan Tsvetkov

Fitch Ratings assessment not inconsistent with what I see. Boise (Ada County) is overvalued 33% as of Q1 2020 based on historical price-income relationship. The most overvalued city in the U.S. at present out of ~800 cities with >50k population.

@Joseph M. I am surprised by the CoreLogic study from 2017. "A market is considered overvalued when home prices are at least 10 percent higher than the long-term, sustainable level." The article does not specify the latter.

As of Q1 2020 based on historical price income-ratios, I see only 11 out of 50 largest cities (by population, not inventory) overvalued > 10%. The rest under- or fairly valued.

This is a study I did myself. Can write to me for any data, further info.

@Jordan Moorhead Overvalued only matters at peak of market cycle, as valuation correlates to the subsequent price drop. During intermediate market cycle prices are driven by momentum, autocorrelation studies confirming.

@Gordon Forbes If asset prices, broadly speaking, are high, portfolio allocation partly shifts towards safe haven ones like gold, maybe in the near future electronic gold etc. To state that those no longer provide protection, need to be able to prove that their negative correlation in crisis times has disappeared (which doesn't seem to be the case now in 2020)

The stock market, absent tech, is not necessarily 'high', but either way you need to have precise valuation metrics to work with: P/E ratios, Tobin's Q, Price/Revenue, Hussman Margin-adjusted CAPE, whichever one correlates best to subsequent price changes. Now tech is hard to 'value', but still need to build a model for that to be able to make any claim.

Real estate outside the U.S. is overvalued in certain countries and regions: Canada, Scandinavia, Oceania, UK some prominent examples according to a study by economist Niraj Shah quoted on VisualCapitalist.

Real estate in the U.S. is overvalued in certain states and counties. Based on valuations I computed as of Q1 2020 using historical price-income relationship Idaho, DC, Nevada, Colorado, Arizona, Florida, Texas and Washington are overvalued >10%. The rest states are either undervalued or fairly valued, though specific counties in them may be overvalued. 

Can PM me for any data, further info.

@Taylor L. Yes, Texas growth in population/job terms is enormous in recent decades. No question about it. 

Now, prices having come up since 3 years ago, does not necessarily mean the market has become more 'overvalued', could in fact be the opposite assuming income growth (or another comparison variable in a different valuation framework) was even higher. It may sound obvious, but important to mention.

@Richard Potts I see both Austin and San Antonio overvalued (at +16% and +12%, respectively, as of Q1 2020). The above based on county-level valuation metrics I compute in-house for my own multifamily investment/data analytics endeavors. Metrics correlated high on the state level vs 2007 actual price drops.

The state of Texas is overall +11% overvalued as of the same time. So yes, Austin/San Antonio definitely overvalued. In fact, Austin is a top 10 overvalued city in America at present. Surpassed only by Plano in the state of Texas.

Thanks Alina for hosting me! Some great feedback.

@Joshua Hittinger What specific analysis lead you to the above conclusion? 

I compute valuation metrics for various states/counties in my data analytics LLC, as well as for multifamily investment out of state.

As of Q1 2020 data it looks as follows: while Washington state is among just 8 overvalued states nationwide (>10% overvalued based on deviation from historical price-income relationship), King County (Seattle-containing) was the least overvalued at ~2% valuation. Spokane, Thurston appear much more overvalued, and then Pierce as well.

So yes, Washington state is overvalued, though King County (Seattle) is the least so in that state. Feel free to PM me if any questions.

@Diogo Marques The key is to devise a valuation metric that would have the highest predictive power vs post-peak actual price drops.

I did a similar study myself for my data analytics LLC, and '% deviation from historical price-income relationship' showed the highest predictive power: 83% correlated at the state level vs. actual price drops post 2007. The above valuation metric is also less sensitive to interest rates, in comparison to cap rates/price-rent.

At the very minimum, need a comparison variable against which the relative valuation is assessed, not achievable if operate just in 'price' space. I can share a video or presentation slides for anyone who may be interested.

Event Details

Stefan Tsvetkov, MS Financial Engineering, will explore the state of valuations in U.S. private residential real estate, by means of internally derived metrics and external vendor reports.

Parallels to financial markets and theory are to be made, and finance risk-return characteristics are to be compared against. Relationship to public and commercial real estate markets is to be touched on.

A deeper dive into specific regions and counties will aim to shed light on individual metrics predictive power. Outlook for real estate post 2020 recession is to be constructed from the data on the basis of 2007 Global Financial Crisis outcomes and current levels of valuation.

Presenter Bio:

Stefan is a financial engineer working on the derivatives trading desk for a major insurance company. Manages USD 90 billion derivatives portfolio jointly with several colleagues.

Three years as an individual real estate investor in the multifamily space across several strategies.

Founder of Envvy Analytics, an investor-geared property analytics software and Pepela Equity Partners, a real estate investment firm. Organizer of Finance Meets Real Estate investor meetup, which is the host of this event.

Event Start: 6:30pm, 18th August 2020

Event Venue: Online, RSVP and link at https://www.meetup.com/Finance-Meets-Real-Estate/events/271484863/

Event Timeline:

6:00-7:00pm - Presentation

7:00-7:30pm - Q&A