How To Build A Balanced Portfolio


Due to the coronavirus scare, global equity markets suffered a 16 percent downswing in just over one week. This illustrates the importance of having a balanced portfolio, or one that avoids systematic biases. In this article, we’ll cover how to build a balanced portfolio and provide various examples of what a balanced portfolio allocation looks like.

The central idea behind a balanced portfolio is how to design an asset management program that gains an efficient strategic exposure to the global markets that requires minimal forecasting or tactical bets on what the right allocation should be.

Most people are biased toward their own domestic stock market and believe diversification mostly involves owning common stock in different companies. The issue is that each asset class has its own environmental bias.

When growth takes a turn for the worse relative to what’s been priced in, stocks can easily lose a year’s worth (or more) of accumulated gains in just a few days. During the coronavirus scare on February 27, 2020, stocks lost nearly 6 percent peak t trough within 24 hours. That’s more than the projected annualized return of US equities going forward.

We believe that holding US equities is likely to generate you a bit over 5 percent in terms of long-run annual returns.

Stocks fell hard and fast when the virus spooked the markets, effectively wiping out 2-3 years of returns in just a few days.

(Source: Trading View)

Due to this environmental bias, stock-heavy portfolios suffer horrendous drawdowns.

This can be seen below comparing the year-on-year returns of an “all stocks” portfolio versus a more balanced allocation, which has rarely has down years and goes through small drawdowns and short underwater periods. (What this balanced portfolio allocation is and how to build a balanced portfolio we’ll get into further down.)

stock market drawdowns

Financial assets are securitizations of cash flows on goods and services. Asset classes are priced based on what an investor would pay, as a lump sum payment, for said future cash flows. The biggest determinate of what makes an asset class move is when the expectation of that income stream changes.

Strategic diversification that performs well throughout the market cycle is achieved through a balance among asset classes whose fundamentals are best suited to different parts of the economic cycle.

The two main macroeconomic factors that influence asset prices are growth and inflation. The trending moves in growth and inflation being above or below expectations is typically what drives trending up and down moves in certain asset classes.

Therefore, shifts in these variables relative to expectations are fundamentally what drives their movements at the very high level – i.e., rising or falling growth and rising or falling inflation.

(i) Rising growth is good for equities, corporate credit, emerging market sovereign debt and EM currencies.

(ii) Falling growth is good for reserve-currency sovereign bonds (both nominal and inflation-linked) and the safest forms of corporate credit.

(iii) Rising inflation is good for inflation-linked bonds, some emerging market assets, and commodities.

(iv) Falling inflation is good for nominal sovereign and corporate bonds and equities.

Having cash on hand is useful for when money and credit are tight and financial assets underperform cash (e.g., 1982, 1994, latter part of 2018).

The counterargument against strategic diversification is commonly:

1. “Why not just take the risk of going ‘all in’ on equities? They have the highest return. The drawdowns are just part of what goes into it.”


2. “Stocks always go up.”

Regarding the first argument, it’s not necessary to go “all in” on a certain asset class when you can diversify among different asset classes and mix them well to produce a more optimized portfolio that can give you the same return for less risk, more return for the same risk, or some combination thereof.

When a portfolio has environmental bias, your expected distribution of outcomes is much wider. Having a balanced portfolio allocation is a form of risk management and you can do this in a way that won’t constrain your long-run returns. (This is the most common knock against diversification.)

Everybody’s favorite asset class – whether that’s stocks, bonds, gold, oil, and so on – is going to drawdown more than 50 percent during their lifetime.

Diversifying within asset classes can help reduce drawdowns. For example, if you are concentrated in tech stocks or energy stocks, you can expect your risk and drawdowns to be worse than having a balanced allocation of equities that has exposure to different “factors” (value, growth, dividends, momentum, size, quality, and so on).

If you had invested in the NASDAQ during the top of the 2000 bubble, you would have lost over 80 percent and been underwater for 15 years. And most investors have a lot of trouble hanging on to investments due to emotions, or suffer the need to raise cash when portfolios erode their collateral cushion, and so on.

In bonds, you have your choice among corporate bonds, safe government bonds, high-yield bonds, different durations, different geographies, and so forth.

Commodities can be thought of as a growth-sensitive asset class and as alternative currencies, and each commodity market is subject to its own supply and demand considerations.

As mentioned, what you can be pretty sure of is that each asset class will act differently. When some assets zig, others zag.

In other words, all asset classes have environments in which they do well and others in which they do poorly.

Consequently, owning the standard portfolio that’s full of equities is equal to taking a large bet that growth will be above the dollar-weighted expectation.

Regarding the second argument, stocks do not always go up, even when thinking out over decades. Stock prices are a function of future earnings discounted back to the present.

So, broadly you need two things to go your way (not necessarily both):

(i) earnings to go up through higher sales and/or margin expansion, or

(ii) lower interest rates to increase the present value of cash flows

Japan’s stock market has been severely underwater since 1989. And not by just a little bit. It’s still nearly 50 percent below its peak.

japan stock market 1989

This is because there was an over-extrapolation of the Japanese economic boom of the 1980s. That didn’t come to fruition and interest rates have been down at zero for more than two decades. When interest rates are effectively maxed out, there is little room to continue to lower them to increase the present values of assets (i.e., their prices).

Bias is a key account killer

There’s a host of reasons why traders get wiped out – e.g., uneducated use of leverage, betting too much on a certain thing, impatience, and the whole subject of risk management gone wrong in its various ways.

But bias is one of the biggest reasons why traders blow huge holes in their account (or worse). Most traders approach the markets with certain preconceived biases, whether its explicit – namely, having certain opinions over what’s going to happen – or implicit (i.e., being skewed toward an equity-heavy allocation as the default).

These biases result in traders holding positions for no other reason that they already own them or believing that exiting them is akin to “giving up” (i.e., a type of status quo bias). This results in making trades that are not based on reality.

The basis behind building a balanced portfolio

The key to having a balanced portfolio allocation is about balancing the risk among different assets.

Stocks are 2x-3x more volatile than bonds.

With the standard 60/40 stocks/bonds portfolio, this is anything but a balanced portfolio because stocks will dominate the movement. Namely, stocks contribute disproportionately to such a portfolio’s risk. It’s actually closer to an 88/12 portfolio in that respect.

A 50/50 stocks/bonds portfolio is closer to a 77/23 even though it’s ostensibly balanced because the dollar amounts are equal.

Moreover, because bonds yield less than stocks, your long-run returns will be lower if you’re trying to equate your risk between the two to achieve balance.

To achieve risk parity between stocks and bonds you’re at something closer to a 30/70 or 35/65 allocation. The problem, of course, is the paltry returns that would result from such a portfolio. That’s why bringing parity between the risk of stocks and the risk of bonds is essential to have an efficient strategic asset allocation.

The key behind a balanced portfolio allocation

Once you have taken the steps to make all asset classes exhibit approximately the same risk, you can begin to diversify for all economic environments without having to forfeit expected returns.

Or, to put it a different way, your search for diversification will no longer need to be constrained by its impact on your long-term returns.

Once bonds are leveraged to match the risk of equities or whatever level of risk you want (e.g., normally this is done through the bond futures market or another leverage technique) one can match the risk contribution of equities or any other asset class without fear of giving up returns.

In other words, in order to achieve balance, the idea is not to excessively diversify away the “growth” part of the portfolio by cutting down its notional allocation, but rather leveraging the fixed-income portion to match the growth portion.

By extension, this type of portfolio will achieve higher return per each unit of risk while also greatly diminishing left-tail risk.

The graph below shows a representation of how various asset classes come as they are in terms of return versus their risk. Different investors will have different expectations relative to what is shown below. But the general point is that investors want to be compensated for taking on more risk.

asset class expected returns

Now if someone wanted to design a portfolio that generated them an expected 12 percent returns per year, they would look at this chart and see that it’s impossible unless they employ leverage.

Moreover, they’d be inclined to buy things toward the upper right of the chart, particularly equities, which are already inherently leveraged (companies have debt). So, they’d have a ton of risk in only one or two asset classes that are relatively correlated to each other.

If they didn’t want that level of risk, they’d have to buy more assets toward the lower left, which would decrease their risk but at the cost of expected returns.

If we wanted to combine these asset classes into a diversified portfolio each returning 12 percent, the trader would need to employ leverage or similar techniques to deliver higher-returning and higher-risk assets necessary to achieve this objective.

They could then be balanced in a way to construct a portfolio yielding the higher return but at substantially less risk if they leveraged only a very limited number of asset classes. Or at lower risk but at a comparable return to equities or their desired benchmark.

A stylized example of this process of leveraging up each asset class and blending them into an optimally structured portfolio is shown below.

how to build a balanced portfolio


Beyond the 60/40 and standard ‘diversified’ portfolios

Balancing your portfolio between the competing forces of growth and inflation requires more than just stocks and bonds.

It is fairly straightforward to balance your portfolio to growth using just equities and bonds, but it is hard to have a balanced portfolio allocation using just those two asset classes. This is because they are both dependent on low interest rates to keep their prices healthy.

If inflation flares up, this increases interest rates. The drop in the price of bonds will depend on duration and credit risk. Higher inflation also brings the prospect of tighter monetary policy. In such environments, you will tend to see both stocks and bonds perform poorly and things like commodities and cash do well.

The US was in this type of inflationary environment in the 1970s. The traditional 60/40 portfolio underperformed. One that was more balanced and had a bit of gold in the portfolio did much better.

Portfolio Allocations

Asset Class %
US Stocks 30%
10-yr Treasury 60%
Gold 10%


US 60/40
Asset Class %
US Stocks 60%
10-yr Treasury 40%


Portfolio Returns

Portfolio performance statistics
Portfolio Initial Balance Final Balance CAGR Stdev Best Year Worst Year Max. Drawdown Sharpe Ratio
Balanced $10,000 $634,639 8.93% 7.27% 31.39% -4.58% -11.94% 0.58
US 60/40 $10,000 $835,204 9.55% 9.94% 31.69% -15.07% -28.54% 0.50

Returns of the balanced portfolio are steadier over time and the drawdowns are not as severe:

balanced vs 60/40


Example Balanced Portfolio Allocation

Take the following example of a more balanced allocation:

Asset Class %
US Stocks 20%
10-yr Treasury 20%
Gold 10%
Cash 20%
Intermediate Treasury 30%


We can diversify further by using different currencies, branching into corporate bonds, municipal bonds, and so on. But to prevent the allocation from getting overly complicated, we’re keeping it to five different broad asset classes.

Compare this to a 100% stocks portfolio:

US Stocks
Asset Class %
US Stocks 100%

Portfolio Returns

Portfolio performance statistics
Portfolio Initial Balance Final Balance CAGR Stdev Best Year Worst Year Max. Drawdown Sharpe Ratio
Portfolio 1 $10,000 $406,188 7.94% 5.24% 25.14% -2.15% -7.99% 0.61
Portfolio 2 $10,000 $1,135,926 10.25% 15.58% 37.82% -37.04% -50.89% 0.41

If the balanced portfolio allocation was leveraged to match the volatility of the “all stocks” allocation, the compounded annualized growth return (CAGR) would be 14.0 percent (versus 10.4 percent for all stocks), with a maximum drawdown of 23.3 percent, versus 50.9 percent for all stocks. (This takes into account the 4.75 percent borrowing rate on cash.)

Measures of return relative to the risk taken on are also superior, and does so at a lower correlation to the market.

The red bars, showing annual equity returns, show the clear environmental bias (and consequent large drawdowns) associated with stock-heavy portfolios.

how to build a portfolio


The role of leverage in building a balanced portfolio

As mentioned, to get asset classes to exhibit the same risk such that you can effectively diversify among all environments, this will involve using leverage in some form. This is fundamentally how a bond can compete with a stock, a piece of real estate, and so forth.

Leverage, when used appropriately as a tool to better accomplish your objectives, can actually lower your risk and/or increase your return to risk ratio.

As mentioned, when designing a balanced portfolio, to get the fixed income volatility in line with that of equities, this is typically achieved through bond futures. This gets you greater access to bonds, with low collateral outlay, to help you balance out your exposure to stocks.

The cash flows of stocks are theoretically perpetual, whereas the cash flows of bonds typically have a fixed end date (i.e., when the bond matures). This is the reason why stocks are generally more volatile overall – they have longer duration. Moreover, in the case of a theoretical bankruptcy, creditors (i.e., bondholders), will be paid before shareholders because they are more senior in the capital structure. So there’s a credit-related risk aspect as well.

Nonetheless, duration is taken as the key determinant of risk premiums. (The basic concept behind duration is, theoretically, if you invested a lump sum, how long would it take you to earn your money back without touching the principal? The longer it takes, the longer the duration.) Under the duration argument, this means that pricing across asset classes would share a common delta.

By extension, one could expect that the volatility regimes among different asset classes to remain relatively proportionate throughout time.

Also by extension, if the volatility among asset classes remains relatively correlated throughout time, this removes the large risk associated with introducing leverage into a portfolio.

To balance stocks and bonds in a portfolio, you would need to leverage the bond portion to match the volatility of stocks. This is done either by borrowing cash and putting it into bonds, or through the leverage technique where you put forward a small bit of collateral for larger exposure to the asset class.

If the volatility of bonds increased by a much greater proportion than stocks over a significant period, this adds risk unless volatility spikes relative proportionally across all included asset classes.

Empirically, changes in the volatilities among different asset classes share a materially positive correlation. Fundamentally, this is because the same set of economic, sentiment, and behavioral factors driving markets has an influence on all of them.

For example, in 2008, the markets went from discounting a higher level of growth in the future to a lower level, which was negative for equities but that same influence flowed into bonds as well, which drove the safest forms of credit (e.g., US Treasuries) higher. The drop in real interest rates also drove forward gold prices, which central banks use as a reserve and institutional investors use as a currency hedge when real interest rates become unacceptably low.

Because this correlation in volatility regimes holds up relatively well, this is good for balanced portfolios because the benefits from the diversification effect are maintained.


How to Build a Balanced Portfolio on Your Own

Broadly speaking, among liquid asset classes, you have your choice among the following:

– Stocks

– Bonds and various forms of fixed income

– Commodities

– Gold (considered on its own, as it fits better as a currency rather than a traditional commodity)

The “growth” part of your portfolio is mostly going to be in the stocks portion. Right now, in developed markets, fixed income doesn’t generate you much yield in either nominal or real terms.

If you use bond futures, your expected return is virtually nil, as it’s not a coupon-bearing instrument. It simply tracks the direction of the underlying. Your yield can come in the form of the “roll yield” – buying a cheaper contract as you “roll” the contract from one maturity to the next, usually quarterly.

How much of what type of bond you buy depends on the duration. If you use short-duration US Treasury futures as your bond component, you’re going to need a larger allocation to balance out your equities exposure relative to if you used longer-duration bonds.

Treasury futures are nominally called 2-years (ZT futures), 5-years (ZF futures), 10-years (ZN futures), 30-years (ZB futures), and ultra-long (UB futures).

However, their durations are often lower than advertised. The “cheapest to deliver” (CTD) bonds are based on finding the security with the maximum implied repo rate. Without going on a tangent about the technical details of what that means, the durations are often different than the one implied by the name.

treasuries balanced portfolio

(Source: CME Group)


For example, let’s say for your stocks allocation, you use ES futures (e-mini S&P 500 futures). This currently contains about $148,000 worth of stocks. The duration of stocks is about 17.4 currently. We can approximate this by taking the value of the S&P 500 index – 2,960 and dividing by the expected earnings twelve months forward, which is currently about 170. This comes to 17.4.

If we multiply $148,000 by 17.4, that comes to 2,575,200. This is an inherently meaningless figure that assigns a certain factor to the expected volatility of the asset.

If you wanted to buy a mix of ZT, ZF, ZN, and ZB futures to balance out this duration, you could find a combination knowing the following:

– ZT: $217,000 per contract * 2 (duration) = 434,000

– ZF: $121,000 per contract * 4.5 = 544,500

– ZN: $131,500 per contract * 7 = 934,500

– ZB: $169,000 per contract * 16 = 2,704,000

This means the following individual combinations would balance out your 1 ES contract, using rounded figures:

– 6 ZT contracts

– 5 ZF contracts

– 3 ZN contracts

– 1 ZB contract

Or some permutation thereof for diversification across maturities:

– 2 ZT, 1 ZF, 1 ZN

Accordingly, your bond duration would approximately match your stocks duration.

Gold is typically best allocated in a 5-10 percent allocation in a portfolio. If stocks are 20 percent of your portfolio, or that $148k, then gold should be some quarter to half of that amount.

Each gold futures contract (GC on NYMEX) is worth about $164,000 currently. MGC futures split that amount into 1/10th of that to help those who don’t quite have a need for such a large allocation. That could mean $49,200 worth of MGC (i.e., 3 MGC contracts, or 3/10ths of a standard GC contract).

And let’s say we want commodities in a 3-5 percent allocation to help with inflation hedging. AIGCI futures give exposure to the Bloomberg Commodities Index. At $7,200 apiece, that could mean 5 contracts, or $36,000 worth.

If we finalize this as our basic allocation this comes to:

Stocks (ES futures): $148k

Short-term bonds (ZT Treasury futures): $434k

Intermediate-term bond (ZF + ZN Treasury futures): $255k

Gold: $49k

Commodities: $36k

Notional total exposure: $922k


In terms of percentage allocations:

Stocks: 16%

ST Treasury bonds: 47%

IT Treasury bonds: 14%

10-year Treasury: 14%

Gold: 5%

Commodities: 4%


Expected Performance

Expected performance can be assessed through backtesting.

However, for various reasons, we do not expect future returns to be like past returns. So backtesting can be misleading for reasons related to the idea that the past is not always like the future.

Investors underestimated the 2008 crisis because such a debt dynamic had never occurred during their lifetimes.

They also underestimated the coronavirus issue because they used the SARS public health scare as a proxy without realizing that the coronavirus was fundamentally different.

Backtesting is useful, but other forms of stress tests should also be run.

If we backtest the above in Portfolio Visualizer, we get the following results compared alongside an all stocks portfolio. Commodities are only part of the data since January 2007, so we have a limited sample:

Portfolio Allocations

Asset Class %
US Stock Market 16%
10-year Treasury 14%
Gold 5%
Short Term Treasury 47%
Intermediate Term Treasury 14%
Commodities 4%


Asset Class Allocation
US Stock Market 100%


Portfolio Returns

Portfolio performance statistics
Portfolio Initial Balance Final Balance CAGR Stdev Best Year Worst Year Max. Drawdown Sharpe Ratio
Balanced $10,000 $17,526 4.24% 3.49% 10.23% -1.27% -6.55% 0.96
Stocks $10,000 $28,761 8.14% 15.91% 33.35% -37.04% -50.89% 0.52

We can see that our return to risk metrics (Sharpe, Sortino) are superior. We only get about half the return, but at just 23 percent of the risk.

Our maximum drawdown was 6.55 percent. If we multiply that by our notional exposure of $922k, that comes to about a $60k loss. Naturally, we should expect to have at least that amount on hand to be able to cover our drawdown.

However, if we do backtest this allocation by lumping the commodities allocation into gold, we can get results all the way back to 1972:

Portfolio Allocations

Asset Class %
US Stock Market 16%
10-year Treasury 14%
Gold 9%
Cash 47%
Intermediate Term Treasury 14%


Asset Class %
US Stock Market 100%


Portfolio Returns

Portfolio performance statistics, 1972 to July 2020
Portfolio Initial Balance Final Balance CAGR Stdev Best Year Worst Year Max. Drawdown Sharpe Ratio
Balanced $10,000 $275,625 7.08% 3.88% 21.21% -0.52% -6.29% 0.60
Stocks $10,000 $1,135,926 10.25% 15.58% 37.82% -37.04% -50.89% 0.41


Here we continue to see much better risk metrics in terms of worst year, maximum drawdown, and super reward to risk metrics to go along with lower market correlation.

Our maximum loss since 1977 has been 7.55 percent. If we call this 8 percent, that means our cash cushion would need to be at least $74k (.08 multiplied by the $922k notional).

What would happen if we were to lump our ZT futures into the “intermediate Treasuries” bucket?

We get the following (and also allows us to test back to 1972):

Portfolio Allocations

Asset Class %
US Stock Market 16%
10-year Treasury 14%
Gold 9%
Intermediate Term Treasury 61%


Asset Class %
US Stock Market 100%


Portfolio Returns

Portfolio performance statistics, 1972 to July 2020
Portfolio Initial Balance Final Balance CAGR Stdev Best Year Worst Year Max. Drawdown Sharpe Ratio
Balanced $10,000 $449,337 8.16% 5.82% 29.16% -3.87% -9.05% 0.59
Stocks $10,000 $1,135,926 10.25% 15.58% 37.82% -37.04% -50.89% 0.41


Annual returns:

risk parity annual returns


Drawdowns, as can be observed below, are kept well under control. Limiting losses is a very important consideration because it carries with it material compounding effects over the long-run. The more you drawdown, the harder it is to climb back out of the hole. If your drawdown is bad enough, it’s almost impossible to recover from it and would essentially require a recapitalization, or essentially starting all over again. Drawdowns of any material size absolutely need to be avoided at all costs.

balanced portfolio drawdowns


Our maximum drawdown over this time was about 9 percent. This means we’d need to have at least $85k, based on nearly 50 years’ worth of data, to be confident enough to withstand a drawdown.


A better way to stress test

Backtesting can be faulty because it uses historical returns.

When future returns will be lower, this will cause future maximum drawdowns to be underestimated.  

For reasons covered in other articles, asset classes have the same risks but lower returns relative to what they’ve had in the past. That means your drawdowns going forward will be disproportionate to your annual gains relative to the past.

To get a forecast of what future maximum drawdowns are likely to be, we can use a Monte Carlo simulation.

Monte Carlo is a type of simulation method for risk analysis that is applicable to a host of different real-world applications. You can either run this in mathematical software like R Studio, or use an online template. Portfolio Visualizer has a Monte Carlo simulation feature built into it and has easy-to-read, interactive graphics, and a host of supplementary data. It’s convenient if you don’t have your own internal models set up.

This allows us to input a starting capital amount, percent allocation, forecasted returns, number of years to simulate, input correlations and volatility, what inflation is likely to approximate (and its volatility), and some other bells and whistles (like how much to contribute and withdraw at regular intervals).

In this case, we’ll forecast returns out 75 years, use historical volatility (a good assumption if their duration hasn’t changed much), and use inflation at 2 percent with plus/minus-2 percent annual volatility.

We can also change up the allocation to a broader set of asset classes because we don’t have to worry about a dataset not going back far enough, as might be the case with backtesting. For example, instead of having all of the stocks allocation in US equities, we can split some of it into emerging markets, which will deliver higher returns over the long-run.

Then it’s a matter of what do we plug in in terms of expected returns?

We also need to have these in nominal terms (i.e., expected returns including inflation).

When we think of the returns to plug in for Treasuries, in this example we are using Treasury bond futures. These are non-coupon bearing instruments. Therefore, the return isn’t the same as the underlying securities.

A general rule of thumb for expected returns is the return of the corresponding cash bonds minus the current cash rate. If the current 3-month yield on US Treasuries is 127bps and the corresponding yield on the “cheapest to deliver” bond underlying 10-year Treasury bond futures is yielding 113bps, your expected return could actually be somewhat negative.

A lot of your return will be built into your “roll” or the process by which you can buy a contract at a cheaper or more expensive price going forward.

For example, the difference between March ’20, June ’20, and September ‘20 ZB futures is about $1,000 per contract. That effectively would make your roll about $2,000 over six months. Off $169k, that’s a roll yield of about 235bps annually.

treasuries roll yield

(Source: Interactive Brokers internal interface)


But using the yield of the underlying bonds minus the current cash rate will give you a good proxy. As of right now, this is about 0 percent for most maturities (and a bit higher for durations above 15 years, though is constantly changing).

Your bond futures will heavily act as a hedge, not a yield-bearing type of investment.

Gold is in the same boat. Over the long-term it yields a little bit better than the rate on cash. But gold is a structurally contango market, which means your roll yield is negative – i.e., you will consistently need to buy a more expensive contract as you roll you position forward. Right now, that’s by a bit more than 2 percent per year. Or about the rate on USD cash plus a small premium. Therefore, using gold futures (e.g., GC or MGC on NYMEX), your net return is expected to be about zero.

If you use cash bonds or physical gold (assuming no storage cost, insurance, and so on), you will see some yield, but futures contracts naturally involve low collateral outlays for a lot of exposure.

Here is ultimately what we entered in terms of the allocations among different asset classes:

Portfolio Allocation

Asset Class %
US Stock Market 13%
10-year Treasury 11%
Gold 13%
Short Term Treasury 36%
Intermediate Term Treasury 11%
Commodities 3%
Emerging Markets 13%

Simulated out 50 years:

Summary Statistics

Summary Statistics, by percentile
10th %ile 25th %ile 50th %ile 75th %ile 90th %ile
Time Weighted Rate of Return (nominal) 0.76% 1.32% 1.92% 2.50% 3.05%
Time Weighted Rate of Return (real) -1.10% -0.56% 0.02% 0.61% 1.15%
Maximum Drawdown -33.57% -28.00% -22.81% -18.84% -16.10%
Safe Withdrawal Rate 1.43% 1.68% 2.00% 2.36% 2.74%
Perpetual Withdrawal Rate 0.00% 0.00% 0.03% 0.61% 1.14%


How much collateral would you need?

In our example balanced allocation portfolio, we can use a mixture of futures contracts:

– ZT (2-year US Treasury bond futures)

– ZN (10-year US Treasury bond futures)

– AIGCI (Bloomberg commodity basket)

– GC (gold futures)

– ES (S&P 500 futures)

– MXEF (emerging market futures).


ZT currently requires $800 in collateral per contract. For two contracts, that’s $1,600.

ZN currently requires $1,740 in collateral per contract. For two contracts, that’s $3,480.

AIGCI currently requires $332.50 in collateral per contract. For two contracts, that’s $1,662.50.

GC currently requires $6,375 in collateral per contract. We’re using one contract.

ES currently requires $9368.50 in collateral. We’re using one contract.

MXEF currently requires $5,000 in collateral. For two contracts, that’s $10,000.


Totaling this up, the aggregate margin requirement of setting up this portfolio is $32,500, rounding to the nearest hundred-dollar amount.

This provides access to a notional exposure of $1.145mm. That also means you need to hold ample amounts of cash on hand.

How much cash do you need?

We’ll have to run our simulation, find our estimated maximum drawdown over a 75-year period, then use that amount (plus perhaps a small amount) to estimate the amount we’d be likely to lose. This will be our required cash buffer.

We can also include the return on this cash amount and add to our expected annual return.



Our 50th percentile return comes to 90bps on this allocation using our assumptions. Our maximum drawdown for this 50th percentile column is 18.9 percent. At the 90th percentile, this comes to a maximum drawdown of 28.2 percent.

Multiplying by our notional exposure, this 90bps of return comes to $10,300 in expected annual return (our return of 0.90 percent multiplied by $1,145k).

To handle a 18.9-28.2 percent drawdown, we’d need a cash buffer of $216k-$323k. Adding in our collateral requirement of $32,500, that’s $249k to $355k.

If you’re generating a one percent return on that cash per year, that’s another $2,500 to $3,500 per year. That gets your expected annual return in the $13k to $14k range off this particular model portfolio.


What’s your percentage return?

This heavily depends on how much cash you use as a buffer.

Despite the low returns of financial assets going forward, you can absolutely earn high returns in the market even without talking alpha risk – i.e., tactical betting on what’s going to be good and bad.

But if you don’t keep enough of a capital cushion to bear the inevitable drawdowns, you’ll inevitably get burned.

Your expected return would be 40 percent on your base collateral – $13,000 divided by $32,500.

If you keep $300k as a cash buffer, within our stipulated range, and earn $3,000 per year off that in safe investments, then your return is:

($13,000 + $3,000) / ($32,500 + $300,000) = 4.8 percent


What if that return isn’t good enough?

Let’s say this percentage return is too low, and you’re willing to take on more risk to get a higher return.

You can do this by allocating more to a higher-returning asset class.

So, let’s say you want to boost the notional equity exposure to 26 percent and more toward emerging markets, where the return will be higher. You are going into a higher-returning asset class at the expense of higher risk.

You will have to determine whether it is worth it to you by running the same type of calculations.

So, now this will be our model portfolio:

– 2 ZT (36% of the portfolio’s notional exposure)

– 2 ZN (22%)

– 5 AIGCI (3%)

– 1 ES (13%)

– 3 MXEF (13%)

– 1 GC (13%)



Now we have something that will return higher. For a notional allocation of what’s now $1.197mm, we have a notional annual return of 151bps.

Our collateral outlay requirement is $37,500.

Those 151bps multiplied by our notional exposure is $20,200 worth of annual return.

The question is then, of course, how much do you need to hold against that to tolerate the downswings?

Now our expected drawdown over 75 years, from the 50th to 90th percentiles, runs from 28.0-40.6 percent. Multiplying this by our notional amount (the $1.197mm), this comes to $335k to $486k.

Because we added more of a more volatile asset class, our drawdowns are larger and have a wider distribution of outcomes.

If we hold $335k in reserve, that means we’ll have a total allocation of around $373k adding in our necessary capital cushion plus the $37,500 collateral requirement.

If you earn 2 percent per year on that $335k reserve through a mix of returns on safe / high credit-quality investments, that’s around $6,700 in extra annual return. Our total return is thus:

($6,700 + $20,200) / ($37,500 + $335,000) = 7.2 percent

The return is indeed higher in this case. But you are dealing with more volatility, larger drawdowns, a higher capital buffer needed, and a lesser probability of earning money in any given year.

Excluding the returns generated on your capital buffer, your expected odds of making money within one year on this allocation (making no tactical adjustments) is about 58 percent. Making money over the course of five years stands at about 72 percent odds.

Expected Annual Return

Expected Annual Return
Percentile 1 Year 3 Years 5 Years 10 Years 15 Years 20 Years 25 Years 30 Years 40 Years 50 Years
10th Percentile -3.93% -0.55% 0.74% 1.97% 2.50% 2.79% 3.00% 3.15% 3.40% 3.58%
25th Percentile -1.39% 2.01% 2.68% 3.36% 3.59% 3.75% 3.88% 3.97% 4.08% 4.18%
50th Percentile 5.07% 4.88% 4.86% 4.87% 4.85% 4.84% 4.84% 4.86% 4.85% 4.85%
75th Percentile 11.32% 7.82% 7.04% 6.39% 6.10% 5.95% 5.84% 5.75% 5.62% 5.54%
90th Percentile 15.76% 10.31% 9.08% 7.81% 7.26% 6.96% 6.74% 6.58% 6.33% 6.18%

Annual Return Probabilities

Annual Return Probabilities
Return 1 Yr 3 Yrs 5 Yrs 10 Yrs 15 Yrs 20 Yrs 25 Yrs 30 Yrs 40 Yrs 50 Yrs
> 0.0% 68.57% 88.07% 93.90% 98.71% 99.60% 99.89% 99.97% 100.00% 100.00% 100.00%
> 2.5% 61.50% 70.34% 76.60% 85.46% 89.98% 92.76% 94.92% 96.24% 98.19% 99.12%
> 5.0% 53.60% 48.82% 48.18% 47.79% 46.90% 46.39% 45.61% 45.39% 44.83% 44.42%
> 7.5% 38.07% 28.83% 21.14% 12.36% 7.92% 5.50% 3.58% 2.36% 1.21% 0.57%
> 10.0% 30.59% 11.92% 5.98% 1.19% 0.28% 0.09% 0.03% 0.00% 0.00% 0.00%
> 12.5% 15.42% 3.71% 0.79% 0.02% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00%


If we include the return generated on cash, our odds of making money in any given year is about 61 percent each year and 75 percent over five years.


What if you want to build a balanced portfolio while targeting a specific dollar return amount?

Then we can reverse engineer how much capital you’ll need to make sure you can earn this amount without having to materially worry about drawdowns.

Remember that with futures contracts, we are not working with dividend or coupon distributions. We are working with pure “capital appreciation” instruments. Returns are also volatile year to year.

If you need a portfolio that can generate you $100,000 in income, you will need a strategic asset allocation mix that gives an expected value of this amount.

But we are dealing with the caprice of financial markets. While you can design a portfolio to target a certain amount of income, a certain level of volatility is inevitable. You can never guarantee you’re going to make money each quarter or even each year even with quality risk controls.

With that said, you will need to identify what your strategic asset allocation is.

(i) How much will each individual asset classes return?

(ii) How much will the entire portfolio earn you?

(iii) To get up to your desired income goal, how much of a drawdown is the mix susceptible to?

(iv) Do you have enough capital on hand to cover that loss?

(v) If not, then you probably can’t achieve that goal at the moment, and will need to design a portfolio that can achieve a realistic return that is in line with the capital you currently have on hand to invest with. Investing is a brick by brick process. It is not a way to get wealthy quickly.


Here’s an example of a balanced portfolio allocation:

– US stocks = 18%

– Emerging market stocks = 10%

– Short-term US Treasuries = 25%

– Intermediate-term US Treasuries = 24%

– Long-term US Treasuries = 12%

– Gold = 9%

– Commodities = 2%


Below is an example of how we could construct this mix to generate $100k per year in income.

Not that this is simply the average annual expectation – remember that we are dealing with distributions, probabilities, and expected values, not certitude over what something is likely to do:

– ES (S&P 500 futures) = 2 contracts

– NQ (NASDAQ futures) = 3 contracts

– RTY (Russell 2000 futures) = 2 contracts

– MXEF (EM futures) = 10 contracts

– ZT (Short-term UST bond futures) = 6 contracts

– ZN (Intermediate-term UST bond futures) = 10 contracts

– ZB* (Long-term UST bond futures) = 4 contracts

– GC (Gold futures) = 3 contracts

– AIGCI (Bloomberg commodity futures) = 15 contracts

(*ZB is likely to generate some nominal return because of the roll yield currently associated with the contract and its positive return over cash.)


Running our simulation, we could expect this mix, at some point, to have about a 25 percent drawdown over a 75-year period as a 50th percentile figure.

Portfolio Allocation

Asset Class %
US Stock Market 18%
Long Term Treasury 12%
Gold 9%
Short Term Treasury 25%
Intermediate Term Treasury 24%
Commodities 2%
Emerging Markets 10%

Summary Statistics

Summary Statistics, by percentile
10th %ile 25th %ile 50th %ile 75th %ile 90th %ile
Time Weighted Rate of Return (nominal) 0.76% 1.26% 1.83% 2.39% 2.90%
Time Weighted Rate of Return (real) -1.12% -0.62% -0.06% 0.51% 1.04%
Maximum Drawdown -31.64% -26.29% -21.38% -17.58% -15.01%
Safe Withdrawal Rate 1.42% 1.65% 1.96% 2.30% 2.65%
Perpetual Withdrawal Rate 0.00% 0.00% 0.00% 0.52% 1.03%

The notional value of this portfolio would be about $5.32 million. That means we would need to hold about $1.3-$1.4 million in reserve.

Our simulation estimates, at the average, about a 1.68 percent return on the notional amount.

That means a return of about $89k per year.

We need about $138k in collateral to set up this position. That means with $1.5 million committed to this portfolio, we’d have a cash reserve of $1.362 million.

If we generate a return of 2 percent on this reserve through a combination of US Treasuries, inflation-protected Treasuries (TIPS), short-duration high-quality corporate credit, that gives us another $27k per year.

From the $89k mean annual expectation from the portfolio and the $27k in interest, that gives us an expected annual value of $116k. That gives us a return of 7.7 percent return on equity ($116k divided by $1.5mm).

If you want a portfolio that generates you $50k per year, you can cut this in half to about $750k in required capital. If you want $30k per year, that bring you down to about $400k needed as a reserve.

You can generate a higher return if you have less of a capital buffer. You could after all set up the “$100k per year” portfolio with just $500k and generate an expected return of 20 percent per year. But it would take only about a 9 percent drop in the portfolio’s notional value to wipe you out. But this is dangerous and not recommended. Essentially you would be overleveraged and you would be inviting something bad to happen at some point.

Many traders and hedge funds don’t survive for very long periods because they use too much leverage. It’s always good to get more with less, but you always need to know what can go wrong and protect against that so it doesn’t happen. One of the biggest issues with beginning and intermediate traders is that they severely underestimate their risk. It hits professionals as well (e.g., Long-Term Capital Management).


How to Build a Balanced Portfolio: Step by Step

1. Identify what your strategic asset allocation mix is going to be.

Namely, how can you get to a “neutral” position such that your volatile “growth” portion (dependent on a good environment) can offset the portion dedicated to capital preservation (that does well in a disinflationary environment where growth runs below expectation)?

How can you get asset classes to exhibit the same risk so that you can diversify for all economic environments without the traditional conundrum of having to forgo returns?

2. Based on that mix, what type of returns are you likely to get?

If your returns don’t look adequate, how can you make adjustments that gives you more upside without taking on unacceptable downside?

Can you improve your reward relative to your risk?

If you can improve your reward to risk ratio, do you have enough on hand to cover your potential loss?

3. What type of drawdowns can you expect based on this mix?

Based on this, how much cash collateral do you need to have to ensure that you can always cover any drawdown? What will you put this cash buffer into – Treasury bills, inflation-protected government securities, a mix of safe government bonds and quality corporate credit?

Whatever you put this cash cushion into, make sure your duration is low to avoid wild swings and that the credit quality is high.

You could just let it sit idle as a cash balance, but your return is likely to be poor.

Cash and liquidity also provides options. When a part of your portfolio does poorly, this gives you the opportunity to pick up quality deals. But don’t go on too much of a spree. You always need enough of a cushion to cover your potential downside. If you don’t, you will probably eventually have a problem.


Tactical asset management is difficult for the average investor to pull off operating in a marketplace full of sophisticated entities with their own information and analytical advantages.

It’s extremely hard for professional investors as well. It is not a made-up statistic that 95 percent of professional investors can’t beat a representative index over enough time.

underperform the market

(Source: S&P Global; Financial Times)


Adding alpha (deviating from beta-related returns) is a zero-sum game. To add alpha, if you win, somebody else must lose. It’s actually negative-sum given spreads and transaction costs.

For this reason, making money in the markets jumping in and out of trades is difficult to do.

The way the average investor is very likely to win over the long-run – if they choose to do it themselves – is by having an asset management program that achieves an efficient, strategic exposure to the global markets that can operate without forecasting or tactical adjustments.

That means they will need to find the appropriate balance of assets that helps them achieve a “neutral” position. Namely, a balanced portfolio allocation that’s not biased toward any particular environment. This means including things beyond equities, such as fixed income, gold, commodities, and gets them access to different geographies and currencies. Then it’s a matter of periodically rebalancing.

The coronavirus episode that saw stocks go down more than 15 percent in a single week illustrates the reality of how fragile the world – and therefore your wealth – can be.

During these periods, you need to ask yourself what kind of diversification you had.

Did your portfolio hold up well in such a down environment?

If not, what can you do to achieve diversification to better immunize yourself from these unexpected events going forward?

The “coronavirus week” that saw stocks plunge every day was bad, but we can have moves that are many times that. We’ve had them in the past. Portfolios should not be exposed to any type of environment.

If you’re overweight a bunch of assets that are designed to operate well in a certain type of environment, you are inevitably going to have very bad drawdowns, no matter what asset class is your favorite (stocks, bonds, gold, etc.).

However, they are largely avoidable. That’s why it’s important to scatter your positions across many markets and asset classes.

The reason why most traders don’t last – both retail and institutional – is because they get over-exposed to a particular environment, have a particular bias, use too much leverage, or a combination of such.

Most traders underestimate the risk they truly have on the table – even if all they’re doing is investing in a standard stock market index fund.

A “diversified” NASDAQ fund lost 80 percent peak to trough when it was “hot” in the late 90s and early 00s. Any bit of leverage on that probably wiped you out.

A “diversified” S&P 500 fund lost 51 percent peak to trough during the financial crisis.

Educated use of moderate amounts of leverage can help you achieve your goals in a more efficient and lower-risk manner. But using too much is like playing Russian roulette. It will inevitably wipe you out. Leverage cuts both ways, so it needs to be used in a smart and moderate way.

Many traders also inappropriately use the past to inform the future. Therefore, they engineer their portfolios off what’s worked in the very recent past even if conditions shift to no longer warrant that view. Many of these strategies haven’t been well stress tested even if they’re supposedly optimized using “artificial intelligence” and other seemingly sophisticated means.

Using the past to inform the future is not always a sensible way to do things when the future is different from the past and new variables enter the system. It might be a good way to build a chess program, or anything that’s a closed system where you know what worked in the past will be an appropriate model to apply to how things will work in the future. But it’s a bad way to approach the financial markets.

You don’t necessarily need to put all your money in a certain strategy. But you need to have something that can help you hold up over the long-term, or give you the upside while keeping your downside within acceptable parameters.

Learning how to build a balanced portfolio and having diversified strategic exposure is key.

Post-coronavirus thinking on having a balance portfolio

Investing in a Zero Interest Rate Environment

Post-Crisis Investing

The Future of Trading & Investing: A ‘Store of Wealth’ Perspective

Stagflation: How It Occurs And Building a ‘Stagflation Portfolio’