Underwriting投资测算

Two questions, one set of assumptions: what this house returns if you hold it as an investment, and how that compares with not buying at all. 一组假设回答两个问题：把这套房当投资持有能拿到什么，以及和干脆不买相比差多少。

Note:注： Southern California here means Los Angeles, Orange, Riverside and San Bernardino counties — the four the data covers. San Diego and Ventura are not in it.本页的「南加州」指 Los Angeles、Orange、Riverside、San Bernardino 四个郡 —— 数据覆盖的就是这四个。不含 San Diego 与 Ventura。

Note:注：

Remaining assumptions (all editable)其余假设（均可编辑）
Held as an Investment当投资持有

Annualised return on the cash you put in. The second line is the cumulative dollar amount. 你投入的那笔资金的年化回报。第二行为累计金额。

Home-price growth \ years held房价年涨幅 \ 持有年限 357
Held as an Investment · Cash Flow当投资持有 · 现金流

Only the assumptions you entered, year by year. Money out is negative, money in is positive. Rent is income here — this is the landlord's book. 只跑你填的这一组假设，逐年列出。出钱为负，进钱为正。租金在这里是收入 —— 这是房东那本账。

Buy versus Rent买入 vs 租住

The buyer lives in the house and collects no rent, so vacancy and management have no effect on this figure. The same money rented instead, with the down payment and every monthly difference invested at the alternative return. The buyer's end value is the net sale proceeds alone: every dollar paid while owning is already counted on the renting side, as the money the renter put into investments instead, so subtracting it here would charge it twice. Positive means buying came out ahead. 这里的买方是自己住，不收租金，所以空置率与管理费率对这张表没有影响。同一笔钱改成租房，首付与每月差额都按替代收益率投资。买方的终值只算卖房净得：他持有期间掏的每一块钱已经记在租方那一侧了（就是租客省下来拿去投资的那笔），在买方这边再减一次就是算两遍。正数表示买房赢了这么多。

Home-price growth \ years held房价年涨幅 \ 持有年限 357
Buy versus Rent · Cash Flow买入 vs 租住 · 现金流

Two blocks: buying and living in it, then renting with the same money invested. Both paths pay in the same amount every year, so the gap row sits at zero until the end — unless rent outgrows that amount, in which case the account runs dry and the rest of the rent comes out of pocket, on its own row. No rent is collected here — this is the owner-occupier's book, so vacancy and management have no effect on it. In the renting block neither the Total column nor any year column is the sum of the rows above it: rent paid and investment gains move money inside the account, and the closing balance is what the account holds rather than a flow, so adding a column counts them twice. 上下两块：买下来自己住，以及租住、同一笔钱拿去投资。两条路每年掏的钱一样多，所以差额行在期末之前是零 —— 除非租金涨过了那笔钱，那时账户会被扣穿，盖不住的租金由租房的人从口袋里补，单独列一行。这里没有租金收入 —— 这是自住那本账，所以空置率与管理费率对它没有影响。租那一块的合计列与每一个年份列都不是上面各行相加：租金支出与投资收益是在账户内部挪钱，而期末余额是账户里剩下多少、不是一笔流量，把任何一列加起来都会把这两样算两遍。

above zero — a gain under “Held as an Investment”, ahead of renting under “Buy versus Rent”大于零 —— 在「当投资持有」里是赚的，在「买入 vs 租住」里是比租房多出来的钱 below zero — a loss under “Held as an Investment”, behind renting under “Buy versus Rent”小于零 —— 在「当投资持有」里是亏的，在「买入 vs 租住」里是比租房少掉的钱 faster than the best entry quarter in this city's record for that column's holding period比这个城市在该列那个持有年限上记录最好的入场季度还快 worse than the worst in that same record比同一段记录里最差的还差 no record for this holding period — this pace was never checked against history这个年限没有历史记录 —— 这个速度没有和历史对过 a dash (—) means there is no annualised return to state — the money put in is gone by the time the period ends, or nothing was put in at all; the colour of those cells follows the cumulative dollar amount on their second line instead破折号（—）表示算不出年化回报 —— 要么投入的本金到期末已经耗尽，要么期初根本没有投入；这种格子的颜色改按它第二行那个累计金额定 Δ change since the last run — percentage points (pp) for returns, dollars for amountsΔ 与上一次运行相比的变化 —— 回报用百分点（pp），金额用美元

How to read this.怎么读这些表。 Figures are before tax — no depreciation, capital gains, or interest deduction. Dollars are nominal at the end of the holding period: not discounted, not inflation-adjusted. The price growth rate comes from the median price per square foot for the whole city, so it carries no allowance for how one particular house differs. The rent growth and cost growth rates are not measurements — they are defaults we filled in, so change them to your own case. Moving costs and renovation are not included. The three anchor rows are growth rates drawn from the record for the holding period in the box above, which is why each one names that period; the grid then puts those same rates through the neighbouring periods as well, where a rate can fall outside what that period's own record holds — which is what the hatching marks. Where a cell is hatched, that pace falls outside everything in this city's record for that holding period. The two directions carry different hatching: stripes one way mean faster than the best entry quarter on record, the other way mean worse than the worst — the legend above shows both, and hovering a cell names the bound it fell outside. Where it is not hatched, the count behind it is a count of overlapping entry quarters, not a probability. Where a percentage is charged on something, its label names what it is charged on, and those bases are not all the same: insurance and maintenance are charged on what you paid rather than on what the house is worth later, while selling costs are charged on the sale price. The remaining percentages are rates rather than charges — the mortgage rate, the growth rates and the alternative return are not levied on any of these figures. Cost growth raises insurance, maintenance and HOA only — property tax follows its own 2% ladder and rent follows rent growth. The mortgage rate is a nominal annual rate charged as one twelfth each month, the way a lender quotes it, so its effective annual cost sits a little above the figure in the box; the alternative investment return is a true annual rate, and so are the annualised returns in the tables, so those two are comparable with each other and the mortgage rate is not comparable with either. A loss can run past the cash put in: the loan is repaid in full out of the sale, so where the sale does not cover it the shortfall is yours — non-recourse protection and a short sale are not modelled. Property tax follows California's assessed-value rule: the base is what you paid. This model raises that base by the full 2% every year — that is the statutory ceiling, not the year-by-year figure. The actual factor is the lesser of 2% and the state inflation index, so in any year it lands under 2% the model charges more tax than you would owe. 所有数字为税前 —— 未计折旧、资本利得与利息抵扣。金额是持有期末的名义美元：不折现、不调通胀。房价涨幅取自整个城市每平尺中位价的变化，因此不含单套房自身差异的余地。租金涨幅与成本涨幅不是测量值，是我们代填的默认值 —— 请按你自己的情况改。搬家与装修未计入。三条锚点行是从上面那个持有年限的记录里取出来的涨幅，所以每一条都标着那个年限；网格再把同样这几个涨幅放到相邻的年限上跑一遍，而一个涨幅在那些年限上可能落在它们各自记录的区间之外 —— 斜纹标的就是这件事。打斜纹的格子表示那个速度落在这个城市该年限的历史区间之外。两个方向的斜纹角度不同：一个方向表示比记录里最好的入场季度还快，另一个方向表示比最差的还差 —— 上面的图例两个都画了出来，把鼠标停在格子上还会说明它越过的那条边界是多少。没打斜纹的，背后那个次数是互相重叠的入场季度计数，不是概率。凡是按某个基数收的百分比，它的标签都写着那个基数，而几个基数并不一样：保险与维修按买入价收，不按房子后来值多少收；卖出成本按卖出价收。其余几个百分比是速率不是收费 —— 利率、几个涨幅与替代收益率都不向上面这些数计收。成本年涨幅只抬保险、维修与 HOA 三项 —— 房产税走它自己那条 2% 的阶梯，租金走租金涨幅。利率是名义年利率、按月取十二分之一计息，和银行报价的方式一致，所以它的实际年成本比框里那个数略高；替代投资收益率是真实年化，表里那些年化回报也是真实年化，所以这两者彼此可比，而利率与它们都不可比。亏损可以超过投入的现金：模型在卖房时把贷款全额还清，所以卖不够的那部分缺口要你自己补 —— 无追索权保护与折价清偿都没有建模。房产税按加州计税基数规则：基数是你的买入价。本模型每年按 2% 足额上调 —— 那是法定上限，不是逐年的实际调整幅度。实际系数取 2% 与州通胀指数中的较小者，因此在系数不足 2% 的年份，模型算出的税高于你实际要交的。

The four tables above answer both questions for this one house. That is the end of the chain: market, valuation, underwriting. 上面四张表把两个问题在这一套房上回答完了。这条链到此走完：市场、估值、投资测算。