ORLANDO, Florida, April 4 (Reuters) - When U.S. government bonds become the epicenter of global market volatility, investors' room for taking on additional risk shrinks, sucking the oxygen out of their risk budget.

That's what happened last month when the U.S. and Swiss banking shocks triggered one of the most powerful rallies ever and days of wild price swings in the two-year Treasury bond - typically badged as a "risk free" asset and one of the safest, most liquid and least volatile securities in the world.

The shock blindsided speculative investors who had been positioned for higher U.S. interest rates and yields. The full extent of the turmoil - hedge funds were among those who got crushed - will become clearer as first-quarter readouts emerge.

Calm appears to have returned as regulatory workouts, emergency buffers and remedial actions from authorities have softened the initial blow from the banking crisis.

But the Treasury market volatility of March 2023 could have lasting consequences as investors may demand higher premiums across the bond yield curve and across the asset spectrum - preferring lighter positions than they would have otherwise.

Investors will now have to factor into their future models a past event that nobody predicted, even though the likelihood of it repeating is equally minimal.

Pim van Vliet, head of quantitative equities at Robeco, notes that investors with shorter-term horizons measured in days, weeks or months will almost certainly see an increase in their risk metrics, such as "Value-at-Risk", or VaR.

VaR measures the potential maximum loss on a portfolio of investments over a given period of time, and is calculated using historical returns, current market conditions, and measures of past and expected volatility.

It is just one risk assessment metric that many investors use to help determine their investment strategies. A pension fund's tolerance for risk, and therefore its VaR, will be lower than a hedge fund's.

"When you plug a two-year Treasury into any risk model now using past returns models, expected risk will be higher going forward," van Vliet said.

WHAT ARE THE CHANCES?

All sorts of investors hold and trade two-year Treasuries for all sorts of reasons: pension funds for the fixed income side of their portfolios; central banks for dollar reserves; banks for collateral and liquidity management purposes; hedge funds for leverage or to short sell.

The volatility last month will have hit investors of all stripes, although the more conservative funds with multi-year horizons, like pension funds, will be able to ride it out.

More speculative investors, like hedge funds and leveraged funds, will feel the pain of the two-year bond's move in the days following the failures of SVB and Signature Bank.

The yield posted its biggest declines since the Black Monday stock market crash of October 1987, and moved 20 basis points or more on seven consecutive trading sessions, the first time this has happened since at least 1976 when comparable data began.

Using data going back to the mid-1980s, markets research and advisory firm Exante Data calculates that the two-year yield's slide on March 13 was a 6.1 standard deviation move, and the one-week fall on March 15 was a 7.3 standard deviation move.

The statistical probability of a 6 standard deviation, or 6-sigma event, is around one in a billion, and the chance of a 7-sigma event occurring in any given day is 0.000000000129%.

This level of risk would barely be acceptable for investors trading a penny stock, crypto token, or distressed emerging market currency, never mind the asset widely considered to be one of the least volatile in the world.

Solomon Tadesse, head of North American Quant Strategies at Societe Generale, stressed that demand and supply dislocations unrelated to the asset itself triggered the volatility shock in the two-year bond, not its intrinsic volatility.

Yet because it happened, investors now have to adjust for the probability, however slim, that it is repeated.

"It will have an impact on the maximum expected loss and/or the probability of expected losses. Those parameters will be impacted by this, because it is an event that will factor into future simulations," he said.

(The opinions expressed here are those of the author, a columnist for Reuters.)

(By Jamie McGeever; Editing by Sharon Singleton)