Lombard loan

The Lombard loan models a debt line as a negative allocation in the portfolio. Its monthly cost compounds interest on a synthetic index: it is not a tradable ETF.

How the leverage works

Add the Lombard asset from search (LOMBARD001) with a negative weight. Other holdings stay positive: net weights must sum to 100%, while gross exposure may exceed it.

Example: 150% World ETF + (−50%) Lombard = 100% net → 50% leverage on your own capital.

Why double the exposure isn't double the return

This is the part that matters, and it's frequently misunderstood. With leverage, asset weights can exceed 100%, but net return does not scale linearly. Two distinct effects are at work.

1. Borrowing cost (explicit)

Every month you pay interest on the borrowed capital, in proportion to the negative Lombard weight (−50% → interest on 50% of equity). The backtester models this by compounding base rate + broker interest monthly on the synthetic Lombard index.

In the Performance charts you can enable the Lombard costs overlay to see this impact scaled onto the portfolio.

2. Volatility drag (implicit in the path)

Leverage amplifies market swings on your own capital. With variable returns, the geometric return (the one that matters for NAV) is lower than the arithmetic mean of the same scenarios: after a −10% you need roughly +11% to recover, and with leverage the swings weigh more on the final result.

It isn't a separate fee: it emerges from monthly compounding of the weighted portfolio (gross assets + negative debt). Even in a market that is flat on average but volatile, a levered portfolio can lose ground against an unlevered one.

In short

Doubling exposure doesn't double expected return. From the gross return on assets you must subtract both the cost of debt and volatility drag.

Simplified example: market +8% average annual, debt cost 4%, leverage 50%. Gross return on assets ≈ 12%; net of debt cost alone (≈ 2%) about 10% remains, but with real volatility the historical CAGR can be lower still, because of volatility drag.

What the backtester shows: NAV already includes both effects in the path. The Lombard costs overlay highlights interest only, not volatility drag; the rolling-returns tooltip carries a reminder about the implicit drag.

Margin call (hypothetical)

On the Performance (NAV) chart, the Margin call toggle (on by default when the portfolio includes Lombard) draws vertical lines in months when a typical broker might have demanded additional margin.

How LTV is computed

$\text{LTV} = \frac{|\text{Lombard weight on NAV}|}{100 + |\text{Lombard weight}|}$

It is the loan / collateral ratio (gross assets), not loan / equity.

Example: Lombard at −50% of NAV → LTV ≈ 33% (50 of debt against 150 of assets). The alert threshold is fixed at 70%: reaching it requires a Lombard weight on NAV around −233%, that is, a very deep drawdown without rebalancing.

The 70% threshold reflects typical maintenance levels on diversified equity collateral. Markers flag only the first month LTV crosses the threshold, and each new crossing after dropping back below 70%.

Historical notice: the marker liquidates nothing and does not modify the backtest NAV. With frequent rebalancing and moderate leverage (e.g. −50%) LTV stays around 33% and often no marker appears; they become relevant with rare or absent rebalancing and deep drawdowns, or with much higher leverage.

The historical reference rate

Additional markup

Additional broker/bank interest (%) (default 0.3%): margin or custody/financing fee applied on top of the base rate.

$\text{effective annual rate} \approx \max(0,\ \text{base rate} + \text{broker interest})$

If Euribor (or the German proxy) is negative and the broker margin doesn't offset it, the rate used stays at 0%: the loan does not generate an interest "credit".

Monthly compounding

Each month the Lombard index multiplies by (1 + annual rate / 12), with the annual rate never below zero.

Settings

The Additional broker/bank interest (%) field (default 0.3%) sits in Settings → Lombard loan in the backtester, and in Settings in FIRE.

Limits

• One Lombard row per portfolio; negative allocation down to −100%.
• Monte Carlo and the Ongaro simulation use net levered NAV; the efficient frontier keeps the Lombard weight fixed and varies only the positive legs.
• Not available in Live Tracker.