patrick.net

Brand Says:

> FAB: Can you comment on some of my observations above
> (2nd from the top)? I am puzzled at how landlords can be
> selling properties that are immediately cashflow positive.
> Can you outline some situations in which that happens for
> medium-sized and large complexes?

If you make a big enough down payment almost any real estate deal will be cashflow positive.

I didn’t see any rent numbers but remember a good rule of thumb to calculate cashflow is:

Annual Rent
- Vacancy & Credit Loss (Usually 5-10% of the Annual Rent)
- Expenses (Usually ~$3,000 + Taxes per Year)
- Capital Costs (Usually ~ $300/unit Year over a 10 year hold for apts and ~$1,000/Year for single family).
- Mortgage cost per Year
= Cashflow

goober Says:

> It’s different in the bay area……seriously!! It’s
> really a different model….
> Everybody makes at least half-a-millon a year
> and no more houses are being built. Median
> home price summer of 2008………….. 5-mil…..

It's funny but I (really) just heard a Realtor tm say basically the same thing (she told me she was worried that I would never be able to buy if I kept renting and was sure that the mortgage meltdown would not lower Bay Area prices at all).

FAB: Did you thank her for her concern?

So if property always goes up in the Bay Area, why aren't realtors telling their sellers to hold their property indefinitely? But I guess it's always a good time to buy or sell a home (tm)!

Well, for many of us here, the really interesting thing about the hike in jumbo mortgage rates is how this will affect the Bay Area, where even stucco $hitboxes require jumbo mortgages.
Skibum,
I am not convinced this has happened. From what I can gather from various blogs (see Socketsite and others), the vaunted Wells Fargo 8% rate for jumbo mortgages is for mortgage broker initiated loans. See the Wells Fargo site for loans originated in house. What we may be seeing is the banks "taking back" the broker's margin and bringing it in house. In a "normal" environment, I think the MBs could make a big stink about it, but given the current state of affairs, the banks can say they are protecting the market from the shoddiness of broker processed loans. Think travel agents and paradigm shifts as we finally emerge from this mess... At any rate, things are tightening up concerning documentation and bigger downpayments, so this should have some effect in the Bay Area.

FAB, I literally meant with a 100% loan, the places appeared profitable. Obviously you can pay cash and calculate ROIC that way. How much does a bank typically require for a downpayment on an income property? Also, what is considered good operating income?

http://www.cohomefinder.com/p/80521/405119.htm

$2,450,000 for 15 units, each 3 bed, 3 bath with W/D. Complex is located one block from Colorado State University. Two bed, one bath units one block away in another condo development are leasing at $800/month. Thus I will estimate these condos at $900-1200/month. I assume that you can actually take his assumable 6% loan on $2,450,000.

So from your calculations above:

Costs: -3000 (expenses) + -13500 (taxes) + -4500 (costs) + -147000 (100% mortgage @ 6%) = -168000/year

0.9 * (12 * 15 * 900) = 145800, loss of 22200
0.9 * (12 * 15 * 1000) = 162000, loss of 6000
0.9 * (12 * 15 * 1100) = 178200, income of 10200
0.9 * (12 * 15 * 1200) = 194400, income of 26400

Assuming that the bank required 10% (245K) or 20% (490K), in the $1100 scenario and subtracting out the mortgage difference, you would be making (10200 + 14700)/245000 = 10.2% ROIC and (10200 + 29400)/490000 = 8.1% ROIC, respectively.

Now, 8.1 % or 10.1% ROIC does not seem like a huge return for the present term. That's basically just treading water in the index funds. And 15-unit apartment complexes require major work to keep running. However, assumedly inflation will push prices upwards, eroding the mortgage deduction.

Now if you could get the place for a significantly reduced capital outlay, say $2M, then the returns start to look very attractive.

DennisN,

Long about 2002, during Harvey Pitt's brief tenure (I believe) they did something that should have been done months prior. They made "positive determination" mandatory. Meaning you couldn't have some whiz kid with a 2k account on e-trade putting in an order to short 2 million shares!

You had to be able to prove you had the resources to deliver said shares in the event you were on the wrong side of the trade. It got to be a real problem b/c there were shorts stacked up well above and beyond the total number of shares outstanding! Too bad it only took two years to get it right?

AFAIK the order book specialist pairs up as many positions as he can and if there are enough longs on the other side of the trade your short will be honored. If not, you're SOL. However... I'm told they have that pretty well policed now. Peter P might be able to fill in some of the finer points.

Randy H,

DennisN had a question (that wasn't dumb at all) and my response is tied up in moderation. A little help?

I think 2 things are happening in the mortgage market, both caused by the "repricing of risk".

1. The banks are putting a larger risk premium on broker-originated loans. I'm not sure this is so much a desire by the banks to bring the business back in house. It could also be a way to force higher standards on the brokers.

2. The banks are repricing conforming loans and therefore being forced to reprice conforming loans at near where jumbos are sitting. Since the market for jumbos is very illiquid right now, no one knows what the fair price of jumbos even are, so the banks are being conservative.

For the record, I've always had a "jumbo" on my BA homes. When I fist bought anything over $250K was a jumbo. That's because the standards are set by national medians, not accounting for regional disparities.

A little help?

Done

Brand

I had to model a very similar scenario (though it was commercial real estate) in B-School. If you run that as a reasonable monte carlo simulation you'll find a lot of loss in the demand and price variability alone.

I have noticed prices are down about 10-15% in LA area.

However there are still knife catchers out there but fewer every year.

Here is an encouraging sign: in 2004 my uneducated neighbor quit his $10 an hour phone job, got a RE lic after 8 months of hard hard study and told me he wuz 'gonna be a millionaire soon'. He starts flipping homes. He did have about 400k equity to start with so a good start. I just saw a pre foreclosure sell for 740k but this guy paid 850k in 2004. He must be broke or near broke as the home was vacant, dead lawn, mail piling up.

flippers dying on the vine every day. This time next year they should mostly be worked out of the system. Im basing this on the high number of flips I still see listed (for 6+months typcially) and the fact that foreclosure takes 12 months and mostly they wont/cant hold on that long. Even rented out some of these places will eat 2k+ a month while rented.(more if ARM loan has adjusted)

http://www.sfgate.com/cgi-bin/article.cgi?f=/n/a/2007/08/07/financial/f140331D05.DTL&tsp=1

Sun Microsystems to Reduce Work Force

What we may be seeing is the banks “taking back” the broker’s margin and bringing it in house.

EBGuy,

On the other hand, there is a piece in bloomberg suggesting that in fact the main reason jumbo loan rates jumped so much is that all of a sudden, there is no demand for them in the secondary market. Most of the demand now is in loans that can be bailed out by Fannie or Freddie if need be, which doesn't include jumbo loans.

I believe this is one componenet of what Randy is talking about with banks, lenders, and investors "repricing" risk. They are overshooting by looking for totally bailout-able loans, imo, but that's what happens during a panic...

Brand Says:

http://www.cohomefinder.com/p/80521/405119.htm

$2,450,000 for 15 units, each 3 bed, 3 bath with W/D. Complex is located one block from Colorado State University. Two bed, one bath units one block away in another condo development are leasing at $800/month. Thus I will estimate these condos at $900-1200/month. I assume that you can actually take his assumable 6% loan on $2,450,000.

-------------------------------------------------------------------------

My goal in this downturn is to pick up a similar deal (1/3 smaller) in here in my state...

1- Long term (10-20 yrs), I am willing to bet 5-to-1 that the risk-adjusted return on similar deal initiated at 2009-2010 prices will be very very attractive....great leverage, credit still relatively cheap, much smaller pool of potential competitors on the available deals (Casey Serin who?), probable high future inflation and high volatility in the stock market...

2- The usual risk-averse suspects who are now talking about 'Monte Carlo simulation, opportunity costs' will be quick/force to drop fancy concepts like ‘ex-ante, ex-post’ in a couple of years, when the attractiveness of such deals will be plain for all to see…Don’t listen to them as they don’t seem to understand very well the concept of decision making under uncertainty and incomplete information (dynamic programming)….

My only concerns are that this deal seems quite big (time/spare cash needed) to pull off with a high probability of success:
1- Managing 15 units on your spare time is a daunting task
2- You will need to have at least a 5-10% rainy day fund (unexpected maintenance...) on top of a 10-20% downpayment/renovation fund...lots of cash...
3- The implicit assumption is that students will rent the place: is that true now and is it likely to change (new dorm, stigma, university shuttle stops nearby..)
4- Can you cope under a worst-case scenario for a couple of years...

Anyway, let us know what you find/decide

Brand Says:

> $2,450,000 for 15 units, each 3 bed, 3 bath with W/D.
> Complex is located one block from Colorado State University.
> Two bed, one bath units one block away in another condo
> development are leasing at $800/month. Thus I will estimate
> these condos at $900-1200/month.

I assume that you can actually take his assumable 6% loan on $2,450,000.

If we assume $1000 a month we get gross income of $180K

Net income after vacancy and collection loss will be ~$171K

Expenses at $3K a unit leaves us with $141K

Backing out taxes and CapX will get you a NOI of $123K (a 5% Cap Rate).

The best commercial loan you will get today will have a rate of about 6.75% and the lowest DSCR you can get is 1.20 so the biggest loan you can get will be about $1,265,000 (52% LTV) so if you buy this property at the list price you will get about a 2% cash on cash return (when the cap rate is lower than the loan rate it kills the cash on cash return). The way I read it you can assume a current loan at 6% but it does not mention the loan amount so it is probably a $1mm loan.

Randy: Is the Monte Carlo simulation goverend mostly by spreadsheet equations, or is it something that I would need software to do? In engineering, we obviously do MC simuations but it takes specialized tools and considerable computing horsepower.

TOS: I am just running the numbers as a mental exercise. I doubt that I would ever buy that place, especially at a price where breakeven was so fragile. I am looking at various opportunities now with a pool of experienced and inexperienced investors (mostly small business opportunities), but I am quite interested in learning how to do the analysis and business case myself.

Brand

You need special software for anything useful. You can find some free Excel-macro driven stuff that is OK for a simple mental exercise. The easiest accurate tools to use are things like Crystal Ball (now an Oracle product) which are Excel plugins.

It's very informative to model some variability into profit models. You can take FAB's model above verbatim, then ask him for some rules of thumb about vacancy rates, deadbeat rates, and rent-price variations. Don't worry about rigorous statistics, just do things like triangular distributions given FAB's heuristics.

The amount of variation that will inject into a profit model for say 10,000 iterations will tell you where the sensitivities are in your financial business model. And it is really the sensitivities that are useful, not the predictions. Predictions are just likelihoods based on randomness, but sensitivity analyses will tell you were your probable points of failure are.

As reference, when I developed The Bubblizer I tested it with Monte Carlo simulation in Excel. The scenario output on the second tab is a simplified version of the results of that simulation.

The Bubblizer shows clearly there is a threshold condition whereby the initial asset book value (the price you pay for your home) overloads all the other conditions and consumes over 95% of the sensitivity in the model. Below that threshold, tax-shelter, mortgage interest rate, and purchase price all have roughly even sensitivities. If you refer back maybe 25 threads to when TOS tried to respond to The Bubblizer (by proxy, she never responded with criticism to it in my forum despite the fact she insisted it was categorically flawed) with her own criticism you'll see where she makes the mistake of assuming that sensitivities remain more or less constant. I had an up on her in the argument, because I'd run the simulations and knew generally where the buy-v-rent equation broke down in absolute terms, regardless of everything else (at a 95% certainty).

Apologies to keep posting about Luminent, but the news on this is just so 'are you fu*kin kidding me?':

This time from the Associated Press:
Just a week ago, Luminent Mortgage Capital assured the market it was not exposed to the liquidity squeeze gripping much of the mortgage industry.

Luminent's markets "have deteriorated significantly and in an unprecedented fashion." Its lenders want their money back. The company suspended its dividend and said it will be late repaying some of its debt.

There is something big going on, if companies like this are getting sucked underwater in just one week. Reminds me of one of those b-grade shark movies where the chick (why is is always a chick?) is splashing around and wham, she's gone and the water starts to turn red.

SP

There's a piece on NPR re: the subprime crisis and credit tightening response:

http://www.npr.org/templates/story/story.php?storyId=12561184

It's all stuff rehashed from other sources, but nice to see it's hitting the left-wing side of the MSM (redundant?)...

Randy: Of course, the interesting thing about a Monte Carlo simulation is that it depends heavily on your initial assumptions. For instance, let's assume that I estimate that college students do a typical range of $200-400 a year in damages (+/-33% in FAB's model), but the maximum is far greater than that (say $1000 for drywall damage, cracked tubs, etc.).

As you point out, the primary MC value is to inform you which variables cause the biggest deltas under which conditions. For a college student rental complex, I'm betting that potential damages and tax changes are by far the two biggest factors. You can count on the university not getting any smaller. You can count on the state population continuing to increase. But you can't count on whether or not the city council will rule against "pre-existing" duplexes made from single family homes, or whether you'll get a bunch of keggers that earn you some fines from the kampus kops.

Off topic, and perhaps this has been discussed before, but I was just thinking about societal trends that might impact housing costs. It has been mentioned that housing costs rose after women really and truly entered the workforce in the 1970's, and suddenly "a little extra spending money" became the norm, so housing prices shot up. My understanding is that was a permanent change.

Which got me thinking about 401k's, the new self-run pension plan. Most first-time buyers seem to dip into a 401k these days--you can take out a $50,000 loan from a 401k without incurring penalties. In fact, they often have special repayment arrangements for home purchases. Add a spouse with a 401k and you're looking at couples with access to up to $100,000 toward a house or whatever. If you don't mind penalties and you have more money in your 401k, well, you could take more. This bounty comes from retirement accounts--but you never could have accessed a pension in the same way. Uncle Sam doesn't give that kind of access to Social Security funds. And 401k's are not counted in the negative savings rate, if I remember correctly.

Do you think that the advent of 401k's as the primary pension format in the US could have had an impact on the formation of this bubble? I remember starting to hear about 401k's in the early 1990's, though they may have been around longer, and it takes awhile to build up a nice retirement nest egg...does anyone here have a sense of when 401k's became prominent enough that just about everybody had one?

I was trying to figure out where those ugly jumbo at 8% rumors started. Looks like it was from this WSJ article. And then CNNMoney keeps repeating it. The reality is that Wells Fargo is not funding Alt-A mortgages originated by brokers, and has only ratcheted up their inhouse jumbo by 1/8% to 7% this past week (unless they are spreading disinformation on their own website). Yes, risk premiums are being adjusted, blood is flowing in the streets, but ARMmagedon has not yet arrived (wait for the IO resets). Looks like banks are trying to restore confidence in the markets by starving the mortgage brokers (is someone being disintermediated?):
“As one visual sign of banks’ cooling to a variety of mortgages they had introduced over the years, the broker’s morning loan rate sheet dropped to one page, versus 10 pages usually. The broker asked not to be identified.”

Randy H Says As reference, when I developed The Bubblizer I tested it with Monte Carlo simulation in Excel. The scenario output on the second tab is a simplified version of the results of that simulation.

The Bubblizer shows clearly there is a threshold condition whereby the initial asset book value (the price you pay for your home) overloads all the other conditions and consumes over 95% of the sensitivity in the model. Below that threshold, tax-shelter, mortgage interest rate, and purchase price all have roughly even sensitivities. If you refer back maybe 25 threads to when TOS tried to respond to The Bubblizer (by proxy, she never responded with criticism to it in my forum despite the fact she insisted it was categorically flawed) with her own criticism you’ll see where she makes the mistake of assuming that sensitivities remain more or less constant. I had an up on her in the argument, because I’d run the simulations and knew generally where the buy-v-rent equation broke down in absolute terms, regardless of everything else (at a 95% certainty).

----------------------------------------------------------------

1- Randy, for a Monte Carlo simulation (Matlab / R / C++ numerical recipes) to make sense, you need to spend a lot of time modeling/understanding the dependencies between the various input variables (correlation structure, joint likelihood and prior distributions, non-stationary effects over time…)

2- It is always the case when you use any of the fancy/complex tools out there, you simply shift the problem but are usually not better off (it is like trying to use a F-18 fighter jet to find the insurgents hiding among the inhabitants of Baghdad)….

3- What I mean is that, in practice modeling/understanding the dependencies between the various input variables is almost impossible without lots of CLEAN NUMERICAL data…

4- If you are not convinced that 99% of your 10000 or so simulations are not at all realistic scenarios (bogus), try modeling the dependencies between inflation, mortgage rates, housing prices, housing cycles, bear/bull markets, probability of bailouts plans/government interventions during presidential elections cycles and irrational behavior of the seller/buyers….

Bottom line:

5- For small financial decisions like small Real Estate investment for small players, you are always better off to use your common sense forecasts and stick to worst-case scenario analyses…

Hi Folks, it has been a while.
Just came back from a long vacation.

I agree with theotherside on the Monte Carlo Stuff.

Well, you can take either a deterministic or a stochastic approach depending on the quality of your data and the model context. The most important input factor is the relationship between your variables. Are they dependent or independent on/from each other ??
Everything else follows on from this initial very important step.

I'm betting that the Fed with sit tight till next year at 5.25%.
I think the Street is playing a game of Chicken with Bernake.

Eliza,

Unless I'm just way, way off here and things have changed radically I "believe" the "first time buyer exemption" is limited to $10,000 and while the 10% early withdrawal penalty is waived, it's still taxed as income.

If people are telling you they're taking out a combined total of 100 grand from their 401k's my guess is that they are grossly misinformed (or more likely... deluding themselves?)

Ahem, in keeping with my theme of "an un-level playing field" this too was enacted in 1997 but we can't dismiss it as a non-factor. With the advent of 100% financing this may have greased home ownership for more than just a few aiding with closing costs. Why not make the exemption applicable for "1st time boat purchases"?

Randy,

Dividends have 2 huge advantages, one of which you alluded to in passing.

1. There is no such thing as a Generally Accepted Accounting Dividend. The company has to shell out cold hard cash, which means (as AHM demonstrated so spectacularly a week or so ago) it has to have some cold hard cash in the first place.

2. You can spend your dividends without reducing your holdings.

There is moral hazard involved with the "growth therefore no dividend" argument. I would rather have a good dividend reinvestment scheme.

(Not investment advice.)

* Notice of disclosure *
I live in Australia, where dividends from tax-paying local companies are NOT double-taxed.

FED is printing money and that is why CountryWide share is going up again (now 28.85):

7-Aug-07 26.30 28.62 25.58 27.35 46,278,100 27.35
6-Aug-07 24.70 26.75 23.64 26.75 50,523,100 26.75
3-Aug-07 25.51 26.20 24.73 25.00 58,082,300 25.00

Just read the Fed’s press release. I guess I’m not as smart as my dog, after all.
What does this part mean to the average Joe?

”Readings on core inflation have improved modestly in recent months. However, a sustained moderation in inflation pressures has yet to be convincingly demonstrated. Moreover, the high level of resource utilization has the potential to sustain those pressures.”

Bottom line:

5- For small financial decisions like small Real Estate investment for small players, you are always better off to use your common sense forecasts and stick to worst-case scenario analyses…

tos, Is this why you failed as an IB/analyst? Were your analyses at work based on your "common sense forecasts" and "worst-case scenario analyses"??

"Mr. Managing Director, here's my report on the financial feasibility of our upcoming merger deal. I know it's only one page long, but based on my own patented "common sense forecast" analysis, I think it's a surefire bet. Yessiree, I sure do. Now when do I get my bonus?"

DinOR,
I think you are confusing Roth IRA's with 401k's.

With a Roth, a first time homebuyer can take up to 10K, you are correct. That is not a loan--that is a withdrawal. There is no 10% penalty, and if I remember right, there is no tax--because the Roth is a post-tax account. The Roth must have existed for 5 years before such a withdrawal occurs.

With a 401k, anyone can take up to 50% of their savings, up to $50,000, as a loan--for any reason. The borrower must pay back the loan with interest. The interest actually goes into the borrower's 401k. There is no tax or penalty, as this is a loan, not a withdrawal. There will be no tax or penalty so long as the loan is repaid.

I had not considered the Roth as well. That means a couple with both Roth and 401K savings could take up to $120,000 out of retirement savings and put it toward a house--50K from each spouse's 401K, plus 10K from each spouse's Roth IRA. Yikes! That could matter, I think.

So are you saying that the Roth bit was enacted in 1997?

TOS

you need to spend a lot of time modeling/understanding the dependencies between the various input variables

You assume that I did not. You are incorrect. In fact, the model is the result of no less than 7 separate individuals' input, at least one of which was actually getting paid to model real estate investment. I simply started the model by adapting a HSBC macro model. From there it was a collaborative work which was reviewed and criticized by probably nearly 100 people. There are actually 3 derivations of The Bubblizer available on my blog as a result of that process, each of which expresses slightly different structural assumptions. All of those models still show that house-price is the overriding sensitivity in the equation.

It is always the case when you use any of the fancy/complex tools out there, you simply shift the problem but are usually not better off (it is like trying to use a F-18 fighter jet to find the insurgents hiding among the inhabitants of Baghdad)….

Cute analogy. But blaming the tools is not valid or sufficient criticism. I used Excel, Crystal Ball and math. The burden is upon you to demonstrate that any of those tools are inappropriate for modeling an investment decision.

Note, this entire model is simply a response to Real Estate Industry folks such as you, who have been running around telling hapless people that they should wontonly leap into a heavily leveraged debt purchase decision because if they don't they're "just covering the costs of a smartly leveraged landlord" as you put it. I and others decided it was time to test your claims and call bullshit. Now you retreat to simplicity as a defense? Perhaps you shouldn't have peddled your snake oil the past couple years in an insane market then. Seriously, I wonder how you sleep at night knowing you've profited off of the financial ruin of others. Have you no sense of decency?

What I mean is that, in practice modeling/understanding the dependencies between the various input variables is almost impossible without lots of CLEAN NUMERICAL data…

We have discrete initial conditions, if that's what you mean by "clean numerical data". We have lots of it, in fact. Since some of those who collaborated on the model are actually professional real estate investors, we have better data than you think we do, and better data than you do. Even humble little me has better data than you realize given that I have a deep in with someone who works for the firm that is even better than yours at selling overpriced condos in prime SF.

If you are not convinced that 99% of your 10000 or so simulations are not at all realistic scenarios (bogus), try modeling the dependencies between inflation, mortgage rates, housing prices, housing cycles, bear/bull markets, probability of bailouts plans/government interventions during presidential elections cycles and irrational behavior of the seller/buyers….

You don't seem to understand the difference between an iteration and a simulation. You also don't seem to understand the notion of modeling only endogenous factors and excluding exogenous ones. I don't have to model a presidential election any more than I have to model the probability of an asteroid hitting the Marina district. Grasping at straws now, aren't we?

Bottom line:

For small financial decisions like small Real Estate investment for small players, you are always better off to use your common sense forecasts and stick to worst-case scenario analyses…

Small? Have you no decency? Are you so depraved that you will do anything to scare, deceive, bully, or lie your way into a commission? You tell any of the clients who've leveraged 98% of their total net worth in this "small" decision. You have a unique talent of sounding intelligent and then discrediting yourself at the very end of your statement. For all but a tiny portion of home buyers, the home purchase decision is the single most critical financial decision they will ever make in their life. I hope they sleep well at night knowing that you were watching their back for them. Try putting your common sense to work on that ethical conundrum.

But hey, the Marina only goes up, right?

Eliza,

Again I believe there is a lot of misinformation circulating about the lunch room. True, you can take up to 50k out (of your VESTED balance) but it's not for "whatever" reason. The IRS is pretty clear here. Education expenses for yourself, spouse or child, to prevent EVICTION from your home, to pay un-reimbursed medical expenses AND a first time residence.

Secondly, not all employers offer loan provisions (particularly smaller employers where the costs are too high) so it's not a guarantee. What's more, the "loan" can't be rolled over and can often become taxable income (as often is the case). With the national avg. 401K balance of 29k I'm going to go on record as saying, it's a bad idea. Was it an influence? You're probably right there. I stand corrected on the amount.

Maybe the bubble was a conspiracy to trap middleclass into home debtorship, thus depriving them the chance to participate in other capital markets.

astrid,

You'll not have been the first to have suggested THAT!

Thanks to Mr. Practical, Kevin Depew, Paul Kasriel, and John Succo for helping explain how a Japanese style deflation could hit the US. The key point is deflation is caused by a massive increase in credit/debt not supported by a commensurate level of income. In simple terms, there is no way to ever pay back what has been borrowed. It's also critical to understand the distinction between a hyperexpansion of credit and a hyperinflationary printing of money. The former is what caused the Great Depression (and is happening again now), while the latter happened in the Weimar republic and is happening right now in Zimbabwe .

http://globaleconomicanalysis.blogspot.com/

The above posting is way too long to repost here. Many posters make their cases for deflation. It is well worth reading.

I favor deflation first and followed by inflation.

DinOR,
While I respect that you know more about many things than I do, the 401K data I quoted is not misinformation. If your 401K has a loan provision--and you are right, not all of them do, but many do--you can typically take a *loan* for half the balance up to 50K for any reason at all, unless the employer has limited the possible uses for the loan--but that would be up to the employer and would require extra work on the part of the employer. I think you are confusing a 401K *loan* (which can be used for just about anything and usually has a longer repayment period for a home purchase) with a 401K *hardship withdrawal* (which the IRS allows only for certain purposes, including a home purchase). See more here:

http://beginnersinvest.about.com/od/401k/a/aa122104a_3.htm

And while the national average 401K balance may be 29K, that is a national average--the numbers for highly paid areas may be higher. Also, the average takes into account boomers who plan to retire on a pension plan, but have put a few thousand into a a 401K because they could. It takes into account the 22-year-old secretary who just started putting $50 a paycheck into her 401K. It takes into account the employee who, after several different jobs, might have retirement funds scattered across a SIMPLE, some IRA's, maybe a 403b, and a 401K. And it also takes into account responsible thirtysomethings who don't believe they will ever see social security and may have more money in their 401K plans than you would think--and may have chosen to spend it on a home now that they are married or the baby is due or whatever.

I don't know. I tend to hang with the responsible thirtysomethings, and many friends who have bought homes in the last few years have taken a 401K loan or withdrawal to do it. It seems to be what people in this age range do. It may be stupid--especially given that the loans have to be repaid within months of leaving the employer who holds the 401K, so taking a loan chains them to the employer unless they want to pay penalties and taxes to the IRS when the loan becomes a withdrawal--but they do it. A lot of down payments are made in this way. And 120K would actually be 20% down on a modest 600K NorCal home.

Eliza,

I may not be as well versed in this particular facet as you might think? Kind of like an MD that specializes in... amputations? Uh, I'd rather -not- think about "that" unless it's a last option?

Also, there are those that would make the case that HCE's (highly compensated employees) actually skew that 29k avg. higher! Moreover, those that were on the much older DBP (defined benefit plan) were mostly from the WWII generation, not boomers. The "hardship/loan debate" aside the truth remains that the loan -must- be repaid! So, while some folks may have taken this path, they'll find they have skirted mort. ins. yet every payday their check feels a little... light?

So now the next step is the "hardship loan"? (Don't laugh, I've seen it happen). You bring up some good points and I'm not discounted them in an off-hand fashion but I fear for the majority of folks it simply represents a willingness to "do anything to get into a house". Lastly, if you had to "tap" your plan for $'s my guess is that they were over extending themselves to begin with. Tell me you're not seriously contemplating this, right?

DinOR,
I assure you, I am not contemplating funding a house with my retirement account. But I do thank you for your concern.

But I think a lot of people do fund houses with retirement accounts, at least in NorCal and maybe all over. And what a lot of people do can impact a market, right? While I would like to see the housing bubble crash next week, I have been trying to imagine whether there might be some changes in behavior that would last beyond the bubble and have even a small lasting impact on housing prices and the way people buy houses. And I can't imagine a reason that people would stop pulling money from retirement accounts in order to pull together a down payment. It's their money, and they can see it and count it and grab it and use it, so my gut says they will keep doing it.

I don't have anything like enough data to say that this 401K/Roth thing will make a difference in the way things settle out after the bubble is done, but it's an interesting idea, isn't it? That, plus the tax-free 250K - 500K profit on the sale of a house may end up mattering even after the credit bubble is history.

Oh, and I probably wasn't clear in my list of examples on the 29K 401K average. I did intend to imply that your highly compensated employees as well as your nervous thirtysomethings probably have a lot more than 29K in their retirement accounts. The confusing list was meant to imply that averages don't tell you much when you are looking at retirement accounts--the age range and the range of situations is too great.