Many people treat the economy as a single cycle and try to guess its top and bottom. That mismatch almost guarantees failure — because the economy does not oscillate in one cycle. It stacks four cycles whose time scales differ by an order of magnitude: a 3-to-5-year inventory cycle, a 7-to-11-year equipment cycle, a 15-to-25-year property cycle, and a 45-to-60-year technology cycle. The layer you operate in decides which layer’s indicators you should watch.
From the Warehouse to the Era
The shortest layer is the Kitchin cycle — inventory — averaging 40 months. Its engine is corporate stock. Demand warms up, orders rise, firms restock deliberately, the economy climbs. Demand peaks and cools, inventories pile up passively, momentum fades. Firms discount to clear stock, destock deliberately, and the economy bottoms. Once inventories are cleared, a new restocking round begins. Watching finished-goods inventory, PMI inventory sub-indices, and short-term commodity swings is enough — this is the core cycle for short-term trading.
One layer up sits the Juglar cycle — equipment capex — averaging 9 years. Machines have fixed depreciation lives; as capacity ages toward its limit, firms invest in expansion and replacement in waves; after overcapacity is released, capital spending contracts until the old capacity is fully scrapped, then a new investment round starts. Track manufacturing fixed-asset investment, equipment purchase growth, capacity utilization, and orders for construction machinery and machine tools. This is the master mid-cycle: one Juglar contains roughly two Kitchin inventory cycles and sets the arc of a full bull-bear market.
Higher still is the Kuznets cycle — construction and property — 15 to 25 years. Birth waves and urbanization create decades-long housing demand; buildings are built and replaced on roughly a 20-year clock; the property chain pulls steel, cement, appliances, and building materials. One Kuznets contains two to three Juglars.
The longest layer is the Kondratiev cycle — the technology-revolution cycle — 45 to 60 years, in four phases: recovery, as a new technology takes its first steps with mild growth; prosperity, as it industrializes fully amid high growth and rising assets; recession, as the dividend peaks, capacity overshoots, and deflation presses; depression, as old capacity is cleared, debts are settled, and the next technology gestates. History has run exactly five: textile and steam, 1782–1845; steel and rail, to 1892; electricity and chemicals, to 1948; automobiles and oil, to 1991; information technology, from 1991 to today.
How to Actually Use the Nest
The nesting is not decoration: one Kondratiev equals roughly three Kuznets, six Juglars, twelve Kitchins. The strongest objection to the framework: the cycle lengths have wide floating ranges, and what certainty is there that four layers turn together — it sounds like astrology. The objection is valid, but it aims at the wrong target. The value of nesting was never to predict turning points; it is positioning — telling you which layer and which phase you are in now. A short-term trader taking direction from the Kondratiev is driving with a map of the wrong scale.
So the framework’s usable conclusion is a single one, and it suffices: align the granularity of your time horizon with the layer you operate in. Trading over weeks, watch inventory data. Betting on an industry’s bull-bear arc, watch equipment investment. Judging the decades-scale direction of property and major asset classes, watch demographics and technology phases. When the granularity of time does not match the layer, even the most beautiful cycle narrative is noise.
