5D Library › Macro-Liquidity Regime Switching, Extreme Tail Dependence and Barbell Asset Allocation · 中文

Chapter 7 Reading the Classics: the “Nine-Year Store” and the Ever-Normal Granary

This chapter reads the earlier discussion of tails and buffers beside a few ancient Chinese texts: a passage on state reserves in the “Royal Regulations” (Wang zhi) of the Book of Rites, the Book of Zhou as quoted in the “Seven Calamities” (Qi huan) of the Mozi, and the ever-normal granary in the “Treatise on Food and Money” of the Book of Han. The aim is not to show that the ancients had mastered extreme value theory or reserve-adequacy ratios, but to see how they expressed the idea of “keeping a buffer for unpredictable bad years”, and to state clearly the boundary between ancient normative writing and modern statistical models. The texts follow Sun Xidan’s Liji jijie, Sun Yirang’s Mozi jiangu and the Zhonghua Book Company critical edition of the Book of Han (孙希旦 撰,沈啸寰、王星贤 点校 1989; 孙诒让 撰,孙启治 点校 2001; 班固 撰,颜师古 注 1962); the “Royal Regulations” has also been compared with the classic text in the Siku quanshu edition of the Liji daquan. The ever-normal granary was discussed in Book 38 and is only recapped here. The English renderings of the classical passages are the author’s translations; James Legge’s translation of the Book of Rites may be consulted for comparison (Legge 1885).

The “Royal Regulations”: “setting expenditure by reckoning over thirty years”

Source text

冢宰制国用,必于岁之杪,五谷皆入然后制国用。用地小大,视年之丰耗。以三十年之通制国用,量入以为出。(孙希旦 撰,沈啸寰、王星贤 点校 1989)

(Book of Rites, “Royal Regulations”)

In plain language: the chief minister (zhongzai, the official in charge of state finances) sets the state’s expenditure at the end of the year, after all the grains have come in. He bases it on the size of the land and on whether the year’s harvest is abundant or poor. He sets the state’s expenditure by reckoning over thirty years, measuring income to determine outgoings.

“Reckoning over thirty years” is generally understood as taking the harvests of thirty years together and using their average to set the scale of spending. The point is that spending should not swing with each year’s abundance or dearth but should rest on the long-run average, “measuring income to determine outgoings”. In modern terms, spending is set by long-run average income rather than current income; the surplus of good years becomes reserves, and bad years draw on them.

The “Royal Regulations”: “a state without nine years’ store is insufficient”

Source text

丧祭,用不足曰暴,有余曰浩。祭,丰年不奢,凶年不俭。国无九年之蓄曰不足,无六年之蓄曰急,无三年之蓄曰国非其国也。三年耕,必有一年之食;九年耕,必有三年之食。以三十年之通,虽有凶旱水溢,民无菜色,然后天子食,日举以乐。(孙希旦 撰,沈啸寰、王星贤 点校 1989)

(Book of Rites, “Royal Regulations”)

In plain language: in funerals and sacrifices, spending too little is called “meanness” and spending too much “extravagance”. Sacrifices are not lavish in good years nor stinted in bad ones. A state without nine years’ store is called “insufficient”; without six years’ store, “in urgency”; without three years’ store, it is said that “the state is no longer a state”. In three years of tilling there must be one year’s food; in nine years of tilling, three years’ food. Reckoning over thirty years, even with disasters of drought and flood the people do not look starved; only then does the Son of Heaven eat, with music at his daily meals.

The words cited in the author’s framework check out against the text; two points need noting. First, some online transcriptions write “凶早水溢” for “凶旱水溢”; the Siku text has 凶旱 (“drought”). Second, this passage belongs to the same section on state expenditure as the previous one but does not follow it directly: between them stand a few sentences on the share of expenditure for sacrifices and for mourning (“祭用数之仂。丧,三年不祭,唯祭天地社稷,为越绋而行事。丧用三年之仂”). “Reckoning over thirty years” occurs twice in the “Royal Regulations”: the first time about how expenditure is set, the second about what the reserves achieve.

Historical context. The “Royal Regulations” is a chapter of the Book of Rites, generally thought to have taken shape in the Han dynasty. Its content is a systematic Confucian design for an ideal royal order (fiefs, ranks and emoluments, land, sacrifices, punishments, care of the aged and so on), not a record of fiscal practice actually followed by any dynasty. The “three years”, “nine years” and “thirty years” are therefore normative standards with strong ritual and political meaning: “the state is no longer a state” is a warning to rulers, and “only then does the Son of Heaven eat, with music at his daily meals” puts the ruler’s enjoyment after the people’s sufficiency.

The arithmetic of “in three years of tilling, one year’s food”

The sentence can be written as simple arithmetic. Let one year’s consumption be 1. For three years of tilling to leave one year’s food, the average harvest must be 4/34/3 a year, a surplus of 1/31/3 a year, which means storing a quarter of each harvest. At that rate nine years leave three, and thirty years leave ten; the three thresholds of “insufficient”, “in urgency” and “no longer a state” (stores of nine, six and three years) correspond to 27, 18 and 9 years of accumulation.

But grain decays. With a fraction dd lost each year, the store follows Sn+1=(1−d)Sn+1/3S_{n+1} = (1-d)S_n + 1/3 and tends to a ceiling of 1/(3d)1/(3d). The program below (the author’s model, illustrative parameters) first computes the deterministic accumulation and then adds random harvests: the harvest varies around 4/34/3 (log standard deviation 0.15), with a 10 % chance each year of a disaster that leaves only 40 % of the usual harvest; a year counts as “looking starved” only when harvest plus store falls short of a year’s consumption.

Calculation

# The arithmetic of the "Royal Regulations" (Wang zhi): "in three years of tilling there
# must be one year's food; in nine years, three years' food". In units of one year's
# consumption c = 1, that needs an average harvest of 4/3 (a surplus of 1/3 a year).
# (1) Deterministic stock with spoilage d: S(n+1) = (1 - d) S(n) + 1/3, ceiling (1/3)/d.
# (2) ILLUSTRATIVE random harvests: lognormal (sd 0.15) around 4/3 with a 10 % chance of a
# disaster year at 40 % of normal; consumption is cut only when harvest + stock < 1.
# 20,000 simulated 30-year spans per policy.
import math, random

def stock_after(n, d, s=0.0):
    for _ in range(n):
        s = (1 - d) * s + 1 / 3
    return s

def span(rng, d, policy, years=30):
    s, short, firsts = 0.0, 0, None
    for y in range(years):
        h = 4 / 3 * math.exp(rng.gauss(-0.01125, 0.15))
        h *= 0.4 if rng.random() < 0.1 else 1
        if policy == "eat the harvest":
            avail, s = min(h, 4 / 3), 0.0                 # no storage: surplus is spent
        else:
            avail = h + (1 - d) * s
        eat = min(1.0, avail)
        if eat < 0.999:
            short += 1; firsts = y + 1 if firsts is None else firsts
        if policy != "eat the harvest":
            s = avail - eat
    return short, firsts, s

if __name__ == "__main__":
    print("years of food in store (deterministic surplus 1/3 a year)")
    print("   spoilage a year   after 3   after 9   after 30   ceiling")
    for d in (0.0, 0.02, 0.05, 0.10):
        cap = "none" if d == 0 else f"{1 / 3 / d:.1f}"
        print(f"   {d:6.0%}           {stock_after(3, d):5.2f}"
              f"    {stock_after(9, d):5.2f}"
              f"     {stock_after(30, d):5.2f}     {cap}")
    print(f"   the ceiling falls below 9 years when spoilage exceeds 1/27 ="
          f" {1 / 27:.1%},")
    print(f"   and below 3 years when it exceeds 1/9 = {1 / 9:.1%}")
    rng = random.Random(54)
    print("random harvests, 30 years: years with less than full rations")
    print("   policy              spoilage  P(any)   mean years   stock at year 30")
    for policy, d in (("eat the harvest", 0.0), ("store", 0.0), ("store", 0.05),
                      ("store", 0.10)):
        runs = [span(rng, d, policy) for _ in range(20_000)]
        anyp = sum(r[0] > 0 for r in runs) / len(runs)
        mean_short = sum(r[0] for r in runs) / len(runs)
        print(f"   {policy:18s}  {d:5.0%}    {anyp:6.1%}     {mean_short:5.2f}"
              f"         {sum(r[2] for r in runs) / len(runs):5.2f}")
    early = [span(rng, 0.05, "store")[1] for _ in range(20_000)]
    hit = [y for y in early if y]
    share = sum(y <= 3 for y in hit) / len(hit)
    print(f"   storage with 5 % spoilage: in {share:.0%} of the spans with a short year,")
    print("   the first one comes in years 1-3, before the stock has been built")
# years of food in store (deterministic surplus 1/3 a year)
#    spoilage a year   after 3   after 9   after 30   ceiling
#        0%            1.00     3.00     10.00     none
#        2%            0.98     2.77      7.58     16.7
#        5%            0.95     2.47      5.24     6.7
#       10%            0.90     2.04      3.19     3.3
#    the ceiling falls below 9 years when spoilage exceeds 1/27 = 3.7%,
#    and below 3 years when it exceeds 1/9 = 11.1%
# random harvests, 30 years: years with less than full rations
#    policy              spoilage  P(any)   mean years   stock at year 30
#    eat the harvest        0%     98.3%      3.88          0.00
#    store                  0%     23.9%      0.31          7.71
#    store                  5%     25.7%      0.34          4.01
#    store                 10%     27.9%      0.39          2.43
#    storage with 5 % spoilage: in 87% of the spans with a short year,
#    the first one comes in years 1-3, before the stock has been built

With no losses, the arithmetic matches the “Royal Regulations” exactly: 1.00 after three years, 3.00 after nine and 10.00 after thirty. With 2 % lost a year, nine years give only 2.77 years’ store and thirty years 7.58; with 5 % the store’s ceiling is only 6.7 years; with 10 % the ceiling is 3.3 years and thirty years accumulate only 3.19. The program finds two thresholds: if more than 1/271/27 (about 3.7 %) is lost each year, a “nine-year store” can never be reached; above 1/91/9 (about 11.1 %), not even a “three-year store”. The nine-year standard therefore implies very low storage losses, or constant rotation of old grain.

With random harvests, a policy of “eating the harvest” almost certainly (98.3 %) meets years of short rations within thirty years, 3.88 years on average; storing the surplus cuts that probability to 24 % to 28 % (losses of 0 % to 10 %) and the average number of short years to 0.3 to 0.4. The remaining risk is concentrated at the start: with 5 % losses, among thirty-year spans with any short year, in 87 % the first one comes within the first three years, before the store has been built. That is what “in three years of tilling there must be one year’s food” means: a buffer takes time to accumulate, and until it has, the system is fragile.

The Book of Zhou in the Mozi’s “Seven Calamities”

Source text

故《周书》曰:「国无三年之食者,国非其国也;家无三年之食者,子非其子也。」此之谓国备。(孙诒让 撰,孙启治 点校 2001)

(Mozi, “Seven Calamities”)

In plain language: therefore the Book of Zhou says: “A state without three years’ food is no longer a state; a family without three years’ food can no longer keep its children.” This is what is called the preparedness of the state.

The “Seven Calamities” lists seven dangers to a state and stresses that “if the granaries hold no reserve of grain, there is nothing to meet famine”. The words it quotes from the Book of Zhou are almost the same as the “Royal Regulations”’ “without three years’ store, the state is no longer a state”, which suggests that “three years’ store” was a common standard in discussions of state reserves from the pre-Qin period to the Han. The Mohist argument stresses frugality and famine preparedness, a different setting from the ritual framework of the “Royal Regulations”, but both treat reserves as a condition of the state’s survival, not merely a fiscal surplus.

The ever-normal granary: stocks cannot be negative

Book 38 has already quoted and discussed the proposal of Geng Shouchang, Deputy Grand Minister of Agriculture under Emperor Xuan, recorded in the Book of Han’s “Treatise on Food and Money”: “buying grain at a raised price when it was cheap, to benefit the farmers, and selling at a reduced price when it was dear; these were called ever-normal granaries. The people found them convenient.” (班固 撰,颜师古 注 1962) Nancy Lee Swann’s annotated translation of the treatise may be consulted (Swann 1950). Only one point relevant to this book is added here. The ever-normal granary differs from the reserves of the “Royal Regulations”: the “Royal Regulations” speaks of accumulating so many years of grain against bad years, the ever-normal granary of buying and selling to stabilise prices. Both face the same constraint: stocks cannot be negative. As Chapter 6 noted, Deaton and Laroque’s storage model shows the consequence of this constraint, price spikes when stocks run out (Deaton and Laroque 1992); Book 38’s buffer-stock program also showed that once the granary is empty, the price ceiling cannot be held. Ever-normal granaries were abolished and revived repeatedly in later dynasties; Will and Wong, with others, studied the Qing state civilian granary system from 1650 to 1850 in detail, including how it operated and the difficulties of managing it (Will and Wong 1991).

Structural comparison and its limits

Ancient text Modern concept Limits of the comparison
“Setting expenditure by reckoning over thirty years, measuring income to determine outgoings” Spending set by long-run average rather than current income The ancient text is a normative claim and gives no method of averaging
“Without nine years’ store, insufficient … without three years’ store, no longer a state” Graded thresholds of reserve adequacy (such as the ARA metric’s 100–150 %) The ancient thresholds are counted in years and come from ritual design, not calibration on a loss distribution
“In three years of tilling, one year’s food” The rate at which a buffer accumulates (the saving rate) The program shows that losses cap accumulation; the text does not consider losses
“Even with disasters of drought and flood the people do not look starved” Keeping a buffer for tail events The text has no notion of probability and cannot be equated with VaR or ES
Ever-normal granary: “buy at a raised price when cheap, sell at a reduced price when dear” Buffer stocks and price stabilisation An administrative and fiscal tool of government, resting on state power; it fails when stocks run out

These comparisons are only structural resemblances. First, the “Royal Regulations” is a design for an ideal order; “three years”, “nine years” and “thirty years” are numbers of ritual and argument, not statistical estimates, and we do not know whether any dynasty held reserves to these standards. Second, the ancient texts have no concept of probability distributions, quantiles or tails, so one cannot say the ancients “had already discovered” extreme value theory or expected shortfall. Third, grain decays and must be rotated, while foreign-exchange reserves carry opportunity costs and exchange-rate risk; their dynamics differ. Fourth, ancient reserves rested on taxation and coercion, and modern adequacy is assessed in entirely different ways. The book therefore says only that the two can illuminate each other structurally, not that the ancients “already discovered” modern reserve theory.

Chapter summary

The “Royal Regulations” of the Book of Rites sets state expenditure by “reckoning over thirty years” and grades state reserves by stores of nine, six and three years; “in three years of tilling there must be one year’s food” means storing a quarter of each harvest. The Book of Zhou, quoted in the Mozi’s “Seven Calamities”, likewise makes “three years’ food” a threshold for a state’s survival. The program shows that accumulating at this rate gives exactly three years’ store after nine years, but that if more than about 3.7 % is lost each year a “nine-year store” can never be reached; with random disasters, storage greatly reduces the risk of short rations, and the remaining risk is concentrated before the buffer has been built. The ever-normal granary faced the constraint that stocks cannot be negative. These are structural resemblances between normative writing and modern buffer models, not forerunners of modern science.

References

Deaton, Angus, and Guy Laroque. 1992. “On the Behaviour of Commodity Prices.” The Review of Economic Studies 59 (1): 1–23. https://doi.org/10.2307/2297923.
Legge, James. 1885. The Li Ki. Vols. 27–28. The Sacred Books of the East. Clarendon Press.
Swann, Nancy Lee. 1950. Food and Money in Ancient China: The Earliest Economic History of China to A.D. 25. Princeton University Press.
Will, Pierre-Étienne, and R. Bin Wong. 1991. Nourish the People: The State Civilian Granary System in China, 1650–1850. Center for Chinese Studies, University of Michigan.
孙希旦 撰,沈啸寰、王星贤 点校. 1989. 礼记集解. 中华书局.
孙诒让 撰,孙启治 点校. 2001. 墨子间诂. 中华书局.
班固 撰,颜师古 注. 1962. 汉书. 中华书局.

Fan, Y. (2026). Macro-Liquidity Regime Switching, Extreme Tail Dependence and Barbell Asset Allocation: Vine Copulas, Hidden Markov Models and Sovereign Buffer Dynamics [宏观流动性体制转换、极值尾部相依与杠铃资产配置] (Version 1.0.0). 5D Library. https://doi.org/10.5281/zenodo.23828049 · Licensed under Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International (CC BY-NC-ND 4.0).